For Albertans, the future of oil and gas is not just another policy debate, it goes much deeper

EDITOR'S NOTE: CBC News and The Road Ahead commissioned thispublic opinion researchin May as the lockdown in Alberta was eased. It follows similar research conducted in March, just as the social and economic shock of COVID-19 was becoming apparent. As with all polls, this one is a snapshot in time.

This analysis is one in a series of articles to come out of this research. You can find links to the previous stories at the bottom of this one. More stories are to follow.

With Alberta in Phase 2 of its economic relaunch and the first wave of COVID-19 seemingly in the rear-view mirror, the Alberta government has signalled that its focus now is to get the economy going again while keeping a lid on the pandemic.

In CBC's Road Ahead poll in May, the survey asked respondents,"What do you think are the most important things Alberta can do to get its economy back on track?" This was an open-ended question, meaning respondents could answer in whatever way they wanted and were not limited to a set of responses. This gives us an unfiltered view of what Albertans are thinking.

The responses show Albertans are divided on the best long-term strategy for strengthening the economy. One group favours economic diversification. Another favours doubling-down on oil and gas.

This might seem like just another policy debate, but underneath lies a deeper rift an existential crisis questioning what it means to be Albertan. Are we about oil or not? Is there a new, post-oil Alberta? Or is it oil and gas or bust?

The intertwining of policy with identity helps explain why politics seemso much more mean-spirited in Alberta right now. These results suggest things could get even messier.

But first, let's look at the data.

What Albertans told us

A quick way to see the responses on the whole is through a word cloud, shown below. This shows the 100 most-mentioned words, excluding commonly-used words like the, and, but, etc. Similar words are grouped, so government, governments, and governing are grouped together under the term govern. The larger a word appears, the more frequently it was mentioned.

Some words clearly stand out, like oil, get, people, gas, workand pipeline. The prevalence of these terms shows Albertans are focused not only on getting people back to work, but also on the energy industry, which has long been the main driver of the province's economy.

That said, there is another term that is also mentioned frequently: diversify. This suggests when Albertans think about oil, many of them are thinking not of continuing Alberta's reliance on it, but diversifying away from it.

This fits in with the results from our March 2020 survey, which found Albertans are split on whether the province should move away from oil and gas.

Diversify vs. double-down

Using computerized text analysis, we can classify responses into four topics based on which words appear in which responses, how many times they appear, and alongside what other words they appear. This is a machine learning technique, so it's not perfect, but the results fit with what we've seen in this and our previous surveys.

Two of topics are short-term actions that address the immediate concerns of keeping the pandemic under control (mentioned by 22 per cent of those who answered the question), so businesses and society can open up, or government supports (19 per cent) to help individuals, businesses, organizationsand social program providers.

The two other topics are more long-term in their outlook: economic diversification (30 per cent), which focuses on increasing the proportion of Alberta's economic activity from industries other than oil and gas, and doubling-down on oil (29 per cent), which calls for not only increasing investment in oil and gas, but protecting Alberta from perceived threats from the federal government and central Canada. These address the structural foundation of Alberta's economy which were problems even before the pandemic hit and will take time to address.

There are two ways to see the polarization in the graph.

One is short-term versus long-term actions, with 59 per cent of Albertans looking at long-term, structural issues with Alberta's economy (diversification or doubling down on oil) and 41 per cent looking at immediate, short-term actions (controlling the pandemic to get the economy going and government supports).

Another way is in left-right terms, with 49 per cent of folks looking at ideas that are arguably associated with the left in Alberta (economic diversification and government supports), and 51 per cent looking at ideas associated with the right (doubling down on oil and getting the economy going).

Polarization even appears in the actual words people used. Consider the following two examples of verbatim, unedited responses that can be classified into the double-down and diversify camps, respectively:

• "Stop transferring money to Ottawa for Quebec and Ontario. We are a have not province since all our federal politicians are doing noting [sic] for oil and gas. Get pipelines built and separate from Canada. Federally JT [Justin Trudeau] obviously does not feel loyalty to Alberta or the west!!"

• "Implement sales tax. Bring back tech company incentives. Bring back energy efficiency incentives. Bring back film industry incentives. Diversify economy. Get rid of corrupt, incompetent ucp [sic] government."

Loyalty. Corrupt. Incompetent. Those are strong words. Not all responses were laden with this much emotion, but many of them were. So the responses don't just reflect policy disagreementsbut also a degree of antipathy toward the other side.

And that's because identities how people see themselves are part and parcel of those policy disagreements.

Policy and identities

This divide in opinion is even more stark when we look at which people are mentioning which topics. The following chart shows the distribution of topics by ideological identification (where someone places themselves on the left-right political spectrum), and their sense of regional versus national identity, which is measured using a question about an individual's sense of relative attachment to Alberta versus Canada.

Those who identify as being on the political left or those who feel more attached to Canada are more likely to mention "economic diversification" than those who identify as being on the political right or those who feel more attached to Alberta. The reverse is true of "double down" responses.

Taken together, thesurvey data shows Alberta is polarized in several ways.

First, there is movement from the centre to the poles in terms of opinions on leaving oil and gas behind. In statistics-speak, this is a shift from a normal (bell-shaped) distribution to a bimodal (valley-shaped) one. Comparing the responses to the question "Alberta should transition away from oil and gas" between our March and May surveys, there is a decrease in the proportion of Albertans who answer "somewhat" agree or disagree and an increase in the proportion of Albertans who answer "strongly" agree or disagree.

It's not just the issue of transition. In terms of how Albertans identify ideologically, there is a shrinking of the centre. This can be seen when comparing data from three surveys the two from this year and the one from March 2018.

Second, Albertans on either side of the issue are affectively polarized, meaning those on one side of the issue harbour animosity toward those on the other side of the issue. This was seen earlier in the two example responses quoted, where both respondents insulted the parties and leaders of the other side. It has also been well documented in the United States, Canadaand other western democracies.

Third, these various cleavages are aligning in a way where they stack and reinforce each other. Each side is becoming more consistent and more homogenous.

Those who identify more strongly with Alberta than Canada, those who identify on the political right, those who oppose energy transition, those who have economically conservative values, and those who vote UCP are becoming one and the same. The same is true for those who identify more strongly with Canada than Alberta, those who identify on the political left, those who support energy transition, those with economically progressive values, and those who do not vote UCP.

Alberta is merely another data point of a trend that can be seen across Canada and the United States.

What's different about Alberta is that energy transition isn't merely a policy questionbut an existential one because oil and gas is a key component of what it means to be Albertan.

And that's a recipe for a fractious and nasty debate.

Is there a way out?

Since the election, the government has accused individuals of being un-Albertan or anti-Albertan when they advocate for, suggestor merely ask about any course of action less than full-steam-ahead for the oil and gas industry.

Never mind for a moment that the government acknowledges that a transition away from fossil fuels is inevitable, at least when talking to external audiences.

The NDP has played into this oil-based narrative by trying and failing to portray Rachel Notley as a champion of the industry. That's not to take away from her centrist track record while in government, but she could not shake the perception that she was only ever a reluctant supporter of oil and gas.

One way to escape this identity-driven polarization is to shift the understanding of Alberta identity away from oil to shared values. This was one of the recommendations of the Alberta Narratives Project, which explored ways to discuss energy transition without furthering polarization.

A similar approach, taken by Calgary author Aritha van Herk, is to understand Alberta identity as a disposition of mavericity a willingness to take risks, look forwardand be creative.

In theory, these approaches are plausible. Previous research we conducted for CBC suggests Albertans do share many of the same values and are not as conservative as the stereotype suggests.

Then again, one of the organizations behind the Alberta Narratives Project was the Pembina Institute the same organization headed by Ed Whittingham, who has beenlabelled as anti-oil and anti-Albertan by the government.

I don't preclude the possibility that regular Albertans can find some common ground with each other and sort out differences amicably. I should also note that polarization in the distribution-of-opinion sense is not necessarily a bad thing. It's affective polarization, or antipathy of the other side, that is dangerous.

Butdecades of political behaviour research has shown that the mass public take cues from elites.

So, there is a role for leaders, politicians, pundits, commentatorsand party activists to model good behaviour and not promote further animosity and us-versus-them thinking.

Data from two surveys are reported in this story. The May 2020CBC Road Ahead survey was conducted between May 25 and June 1, 2020, by Edmonton-based Trend Research under the direction of Janet Brown Opinion Research. The survey sampled 900 respondents, randomly selected from Trend Research's online panel of more than 30,000 Albertans. The sample is representative of regional, age and gender proportions in Alberta. A comparable margin of error for a study with a probabilistic sample of this size would be plus or minus 3.3 percentage points, 19 times out of 20. For subsets, the margin of error is larger.

The March 2020 survey was conducted between March 2 and March 18, 2020, by Edmonton-based Trend Research under the direction of Janet Brown Opinion Research using a hybrid methodology. This survey sampled 1,200 Albertans, randomly selected by phone, and representative along regional, age and gender factors. The margin of error was plus or minus 2.8 percentage points, 19 times out of 20. For subsets, the margin of error is larger. The March 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. Trend Research contacted people using a random list of numbers, consisting of half land lines and half cellphone numbers. Telephone numbers were dialled up to five times at five different times of day before another telephone number was added to the sample. The response rate among valid numbers (i.e. residential and personal) was 13.2 per cent.