The Polls, the Press, and All the Ways the Information Loop Goes Wrong

The Polling frenzy 

In the wake of the Democratic Party changing candidates there has been a frenzy of polls and press releases on the polls in anticipation of the November presidential election.  It is a fitting moment to reflect on the symbiotic relationship between the polls and the press, and how this relationship, this information loop, manages to go wrong.  My first observation, based on a decade of writing this blog, is that the typical scenario is for the legacy media to present a particular narrative--China is corrupt, Russia is evil, Trump is a dangerous clown (to begin with the most obvious examples)--and then to report, with a tone of seriousness and/or surprise, that a recent poll shows a majority of the population in agreement with the established narrative.  The question I always ask is "How could the majority of Canadians and/or Americans possibly think otherwise after they have been submitted to a brainwashing level of the established narrative provided by the media that most of us turn to for our news?"

It's So tempting to believe the numbers

I understand the allure of quantifications of people's feelings and attitudes.  How reassuring to know how the majority of Canadians feel about this or that issue.  And fascinating to know--with numeric specificity--how divorced white men feel about this, and what teenage girls think about that, what the attitudes of the majority of married black women are.  But can and should we believe the polls?  I suspect that the media (and not just politicians) view the polls as evidence that their messages are getting across.  In the other direction, I am also suspicious that pollsters will be tempted to produce polling results that the press will be interested in publishing.  Surveys are very easy to manipulate and corrupt, as demonstrated by a brilliant episode of the satiric comedy Yes, Prime Minister:

Sometimes the polls are just wrong

Sometimes the polls are just wrong to begin with.  When Prime Minister Justin Trudeau refused to release the Huawei CFO from house arrest in exchange for the "two Michaels" imprisoned in China, the Angus Reid poll claimed that 72% of Canadians agreed with him.  Given the recent flurry of prisoner exchanges between the USA and Russia, and the fact that we were holding CFO Meng because of an American extradition request, I wonder how many Canadians still agree with the PM's naive claims that prisoner exchanges are always immoral and dangerous. (See Where's the Canadian Outrage?)  

And the survey is the misinformation

Timing aside, the pollsters got the question completely wrong, then invited Canadians to join in their misinformation.  The survey asked if we Canadians should follow the law and let the "independent" judiciary decide the Meng extradition, or should we break the law and have the government intervene.  Of course, 72% of Canadians said that we should follow the law.  Except, of course, the law in this case--the Canadian Extradition Act--specifies the opposite.  Letting the judiciary decide (if this were possible) would be breaking the law.  Government intervention would be following the law.  (See What Have We Learned from the Catastrofarce?) This is how the information loop functions (or rather, mis-functions).  The survey only proves that the misinformation provided by the government and the media, and reinforced by the survey itself, was getting through to the Canadian public.

Who Did the survey matters

Perhaps the most important question to be asked about a survey, as implied in the Yes, Prime Minister satire, is “who did the survey?”  In October 2022, the Globe and Mail reported that

An August poll by the Taiwanese Public Opinion Foundation, an independent and non-partisan organization, found that only 11.8 per cent of respondents favoured "unification" with China.  Fifty per cent of those surveyed said they would opt for independence, and 25.7 backed the status quo.  

Except it turns out that the “independent and non-partisan” Taiwanese Public Opinion Foundation has its head office in Washington, D.C.  The Board and Staff of its parent organization, the Global Taiwan Institute, have significant ties with the US Defence and/or State Departments, and describe themselves as sharing “a passion for greater ties between Taiwan and the USA.”  (See Survey Says . . . .)  

Sometimes journalists can't seem to read

The weirdest way that the information loop goes awry is that some journalists fail to or choose not to accurately quote the survey they are reporting.  In an article entitled “What Putin Really Wants,” published in Quillett, Christopher Miller claimed that "The vast majority of Ukrainians reject them [the Minsk Accords]."  The Minsk Protocol was an agreement between Russia and the Ukraine in 2014 overseen by the OSCE and mediated by France and Germany to end the fighting between East and West Ukraine by granting increased autonomy to the eastern regions.  Miller's source for the claim that the vast majority of Ukrainians reject the agreements is an article in Euromaidan Press:  "Three-fourths of Ukrainians oppose Minsk accords in current form, poll shows."

When Reporting:  Words matter

The key phrase in the headline is “in current form.” What the Euromaidan article and underlying survey reveal is that only 11% of the respondents claimed to be familiar with the content of the accords.  More importantly, the respondents did not “reject” the accords but, as the article highlights in bold, “the majority of Ukrainians (54%) believe that the Minsk accords should be revised.”   If anything, the survey showed that, far from rejecting the accords,  the respondents wanted the accords in some form but differed on the format.  Even the claim  that the poll represented  “the majority of Ukrainians” is disingenuous, since all 2500 respondents to the survey were from western Ukraine and Ukrainians from the east (Donbas and Crimea) were not surveyed. (See On Blaming America for Russian Aggression.)

Conclusion

We may all be familiar with the cliche that “statistics can be made to say anything,” but surely it is an awful exaggeration that the information loop between the polls and the press can go so wrong on some of the most important issues of the day.