Why Canadian politics is better than American politics.
Politics in the United States has become a place for grifters to make money, while Canada has quality control and focuses on integrity.
Before I continue, if you are interested in joining our Canadian Substack, The 11th Province, which is an Pro-Canadian American perspective on Canadian politics, please click below to Subscribe:
For years I’ve been banging on about how bad American politics has become. And when it comes to the world of polling, and particularly polling aggregates, it’s become even worse.
During the last election cycle, the top polling aggregators (Nate Silver, RealClear, and FiveThirtyEight) allowed polls that weren’t just dubious, but outright absurd. The would have no polling history, hid who their staff were, didn’t provide any crosstabs, or were proven to have just made up data when looking at the unweighted crosstabs specifically.
Here are some examples:
Patriot Polling - started by two high school students with no experience in public opinion or survey research. They recently released a poll saying that most Greenlanders wanted to join the United States.
Trafalgar Group - uses a “Social Desirability Bias” method, which totally takes out random sampling, and instead inserts their own opinion into what should be a random, scientific method.
SoCal Strategies - started by someone who STILL hasn’t revealed their identity, and basically popped up during the election, usually to push a Republican narrative.
AtlasIntel - a Brazilian polling firm that has been deemed “accurate” according to Nate Silver and the old FiveThirtyEight website, but have extremely inaccurate crosstabs, showing that Trump was going to win 47% of black voters in Pennsylvania, and that Harris was performing better with male voters while female voters supported Trump.
AtlasIntel is the perfect example of what is wrong with American politics. I did a video about AtlasIntel on my YouTube channel (click here for the video) talking about the discrepancies in their crosstabs. However, both Nate Silver and FiveThirtyEight have deemed them “accurate” because they were the closest to the horserace number (Trump vs. Harris).
Of course, there are many problems with this. The first and obvious problem is that the crosstabs show that their overall horserace number wasn’t based on a high-quality sample. Essentially, they got it so wrong on so many levels that it ended up being correct somehow. The second problem is that the aggregators purely look at the final numbers instead of what the underlying data says. This creates a fallacy of thinking that these polls are good, when they simple were lucky enough to hit the bullseye twice. They are the polling equivalent to Timothy Dexter.
Essentially, most of the polls that are presented on the “aggregator” websites are garbage, and the underlying data purely shows this to be the case.
Now, let’s go over to Canada.
A few days ago, a new poll popped up from someone called “Kolosowski Strategies”, started by a guy who is a trustee on the school board in York. This poll, which was the only poll that showed the Conservative Party ahead in Canada, was initially included in the Wikipedia list of polls (but has since been removed).
Furthermore, the website for this so-called polling firm was just created a few weeks ago, and the phone number on this site connects to a pest control website, which seems to be owned by Robert Kolosowski, the namesake of this “polling firm”.
I reached out to him to ask him about his firm, and to get some clarity about the relatively short history of his company. He responded, saying that he had just started it, but had experience in survey research.
But there were a lot of problems with his poll. First, he was weighing the poll based on 2021 votes. As many who observe Canadian politics know, the Conservatives won the plurality of the popular vote in 2021, though the Liberals won the most seats. The poll being “weighed” in this ways would lead to the Conservatives being ahead almost every time.
Another problem was with his sample in general. According to the report he published (which you can see here), he sampled 1,448 respondents. Of those, 870 indicated that they voted for the Conservative Party in the last election. That would mean that 60% of his respondents voted for the Tories, when only 34% of the popular vote in 2021 was for the Conservatives. In addition, oversampling of a certain group is usually done to get a better understanding of a smaller subgroup that you want to include in your overall sample. The fact that the Conservatives had the most votes in the last election means that they should have been the least likely group to be oversampled.
The second issue that only 22 respondents in his poll said they voted for the Bloc Quebecois in the 2021 election, and 28 for the NDP. This means only 3.5% of his sample were Bloc/NDP voters, yet they consisted of 25.5% of the overall 2021 electorate!
Furthermore on the NDP, he said that his poll was able to show “one third of NDP voters in 2021 were moving over to the Liberals in 2025”. However, having only 28 respondents would not nearly be enough to glean anything regarding NDP voters in this election, as that small of a sample size would absolutely be statistically insignificant (p-value more than .05).
In addition to that, he had more people sampled who votes for the People’s Party than the NDP and Bloc Quebecois COMBINED!
So, I decided to email him again asking about some of these issues. Kolosowski responded to my first email in 30 minutes. It’s now been two days and he hasn’t responded to my second email (though I know he opened my email because I have an email opening tracker). You can see my response email below (please forgive the typos).
If this poll had been conducted in the United States, I’m almost certain that it would have been included in the aggregates conducted by Nate Silver, RealClear Politics, and the former FiveThirtyEight, because of the low bar that have set. Basically, anyone is admitted.
But in Canada, that is not the case.
This is where our protagonists Éric Grenier and Philippe Fournier come in. These two men are the top polling aggregators in Canada. They could have easily included the Kolosowski Strategies poll if they wanted to, but they have refused. I contacted Mr. Fournier about why he hasn’t included them, and he stated that he didn’t know who this was and that Kolosowski hadn’t reached out to him. I assume it’s the same reason Grenier hasn’t included them in the CBC Poll Tracker.
And this is the reason why I have almost exclusively written about Canadian politics lately, and have become absolutely repulsed by American politics. In Canada, there are people who want to know the truth, and to provide the best academic and journalistic quality when it comes to political number crunching. Only polls that are deemed are credible are considered seriously. Both Grenier and Fournier have shown us what quality looks like, even though I’ve been critical of Grenier’s methods in the past.
In the United States, subpar polls conducted by dubious “companies” (if they can even be called that) are allowed. Hopefully, this frivolous cancer will not spread to the Great White North.