CBC poll: Results give us an idea of who the vaccine hesitant in Alberta really are

This column is an opinion from data scientist John Santos. For more information about CBC's Opinion section, please see the FAQ.

The day Alberta expanded vaccine eligibility to those over 30 years of age, more than 100,000 individuals signed up. This is much-needed good news in the province with the highest per-capita COVID-19 case counts in North America. And yet, the latest CBC Calgary: The Road Ahead survey shows 20 per cent of Albertans have adopted a wait-and-see approach to vaccination, with another 14 per cent saying they refuse to get vaccinated outright. 

While it is difficult to explain exactly why any one individual is vaccine hesitant, there are clear patterns in terms of who is vaccine hesitant. And, those patterns show how politicized the COVID-19 pandemic has become. 

But before we look at those trends, an important point about methodology is in order. Most survey companies report a series of two-by-two tables (commonly called crosstabs or pivot tables) to show the relationship between one variable and another. 

This type of analysis often reveals many interesting patterns. But, it has difficulty isolating which of those patterns are the "true pattern" (the signal) and which are spurious (noise). 

Separating the signal from the noise

For example, anti-COVID-restriction sentiment and vaccine hesitancy seem to be more prominent in rural Alberta. Elsewhere, political attitudes have been linked to vaccine hesitancy in both academic research and in the press. The problem is rural residents tend to be more conservative than city folk, so which has more bearing on one's vaccine attitudes — where they live or what their values are? 

To deal with that issue, I use a method called regression analysis that allows us to look at the relationship between an outcome of interest (vaccine hesitancy) and a factor that might be predictive of that outcome (region, education, vote intention, etc.) while controlling for the effect of all other factors.

This figure below shows the results of a model predicting vaccine hesitancy using region, gender, age, education, income, employment status, having received a COVID-19 benefit, economic conservatism, populism, self-placement on the left-right political spectrum, and reported vote intention.

The most important factor is populism, which we measure using a scale comprised of four categories: trusting down-to-earth people over experts; preferring strong leadership over debate and deliberation; support for increased use of referendums and plebiscites; and believing politicians soon lose touch with the people after they are elected. 

Someone who strongly agrees with every one of these has a 50/50 chance of being vaccine hesitant whereas someone who strongly disagrees with each has only an eight per cent chance. Another way to understand this is that the most-populist Albertans are, on average, 6.25 times more likely to be vaccine hesitant than the least-populist Albertans, all else being equal. 

Where someone sees themselves on the left-right political spectrum matters, too. The most right-wing Albertans have a 43 per cent chance of being vaccine-hesitant whereas the most left-wing Albertans have only a 17 per cent chance.

Education matters

Another interesting finding is that education matters more than vote choice. By this, I mean the difference in probability of being vaccine hesitant between those with the highest and lowest levels of education (21 percentage points) is greater than the difference in probability between supporters of various parties (about 10 percentage points). 

Finally, while not shown here, there is no relationship between having economically conservative values and being vaccine hesitant once we control for other factors. By "economic conservatism" I mean a general preference for lower levels of market regulation and government intervention. 

We measure this using two questions: whether one believes job creation should be left to the private sector and whether they believe in trickle-down economics. While those who strongly agree with both are five percentage points more likely to be vaccine hesitant than those who strongly disagree with both (33 per cent versus 28 per cent), we cannot be confident this difference is statistically distinguishable from zero.

So, what does this mean?

These results show a subtlety in the relationship between political orientation and vaccine attitudes that is often lost in public discourse.

Media outlets have spilled much ink about the higher rates of vaccine hesitancy and COVID skepticism on the right. But while those empirical patterns are readily observable, what this more-nuanced analysis suggests is that it is not "conservatism" in its economically focused form that is driving vaccine resistance. If that were the case, we would not see prominent conservatives posting photos of themselves receiving their COVID-19 vaccine.

Rather, the factors that are actually important include populism (which can occur on both the left and right side of the political spectrum) and left-right ideology in an abstract or symbolic sense. This form of ideology has less to do with the substance of what one believes in and more to do with where one sees themselves and others "fitting" in the political world.

These findings echo earlier research that overturned popular assumptions that, pre-COVID, those who were anxious about vaccines were a bunch of left-leaning hippies. And, there is a broader set of research that shows biased processing of information is just as prevalent on the left as it is on the right. 

If vaccines are the way out of the pandemic, then we as a society will need to convince as many people as possible that they are safe and effective. The good news is that vaccines are actually an issue where there is much common ground between the left and the right. But those on the fence between acceptance and hesitancy — many of whom are likely conservatives of a moderately populist disposition — might need a gentle nudge to get them to do the right thing for the collective good. 

The full-blown vaccine refusers are probably impossible to convince. But, it might be possible to reach the fence-sitters, if they are not demonized by exasperated progressives and stereotyped and lumped in with the vaccine-refusing (and COVID-denying) ultra-populists.

This CBC News random survey of 1,200 Albertans was conducted between March 15 and April 10, 2021, by Edmonton-based Trend Research under the direction of Janet Brown Opinion Research. The sample is representative along regional, age and gender factors. The margin of error is +/-2.8 percentage points, 19 times out of 20. For subsets, the margin of error is larger.

The survey used a hybrid methodology that involved contacting survey respondents by telephone and giving them the option of completing the survey at that time, at another more convenient time or receiving an email link and completing the survey online.