Members of Canada’s political class should on balance receive at least a passing grade for their management of the pandemic, and Canadians should be both proud and worried about policymakers’ performance during the COVID-19 pandemic. The three areas in which they could have done better include: keeping schools open, using technology to combat the pandemic, and prioritizing vaccine targeting.

By now, it seems clear that Canada will come out of the COVID-19 pandemic with a solid all-round performance, but one that was far from perfect. Indeed, if we consider only the Anglo-American democracies, we see that Australia and New Zealand have experienced almost no deaths (36 and five cumulative deaths per million, respectively), while the United States and the United Kingdom have experienced a remarkably higher number (1,830 and 1,892 cumulative deaths per million, respectively). Canada falls right in the middle, at 697 per million. On vaccination uptake, the story is different still, with Canada and the United Kingdom at the head of the pack, the United States lagging (despite ample supply) and Australia far behind. There will, of course, be ample time to take account of Canada’s performance and allocate blame and credit for the actions that were taken by various levels of government, in Canada and abroad. The purpose of this report is not to do that, but instead to ask what could have been done differently in Canada. Specifically, I seek to look back and find alternative paths Canada could have taken in its now 18-month campaign against the coronavirus.

The purpose of this retrospection is not to be unduly critical. Certainly, members of Canada’s political class should on balance receive at least a passing grade for their management of the pandemic. They have been motivated by a desire to protect Canadians, have consistently sought to follow the advice — sometimes contradictory advice — of experts, and have for the overwhelming part governed themselves not only by the letter, but also by the spirit, of the laws and guidelines they have enacted. Instead, my goal is to understand whether different courses of action were possible at three key policy junctures. From the perspective of what citizens would consent to or agree to, at least, it does seem possible that other policy paths could have been taken.

I look at three different sets of policy choices in particular. First, I examine the application of lockdowns, in particular the trade off of locking-down various essential and non-essential services in exchange for keeping schools open. Second, I look at the use of technology-based tracking and tracing to enable smarter lockdowns. Third, I consider the prioritization of vaccine roll outs.

The analysis that follows turns on an important question or counterfactual: What other policy choices or approaches would citizens have agreed to, if politicians had been willing to lead them there? The assumption here should be clear: Citizens have looked to politicians to set the policy frameworks for getting through the pandemic. They have been unusually deferential, non-polarized and non-partisan (Merkley et al. 2000). Especially among citizens who trust in public expertise, they have closely aligned their own behaviours with official recommendations (Merkley and Loewen 2021). Citizens have taken the paths laid out for them by politicians.

What the report finds is that at all three of these stages, different paths could have been taken. These are paths that would have prioritized worse-off Canadians more and that arguably could have freed us out of the pandemic faster and with less economic and social cost.

In the weeks and months following the onset of the COVID-19 pandemic, governments across Canada demonstrated a remarkable regulatory capacity. In the early days especially, this took the form of the closing of whole sectors.

While different provinces took different approaches, and some provinces had substantial variation within their borders, the basic model was quickly set. Some sectors were closed completely while others remained at least partially open.

Across the country, however, schools and other educational institutions shuttered their doors, sending millions of kids home to learn from screens. While partial reopening happened through the autumn of 2020, most students found themselves returned home when a third wave rolled in in the winter of 2021.

While the term “lockdown” bundles all of these policies together in a somewhat unhelpful way, it does describe well the underlying policy goal. By limiting activities, governments limited contact between people, thus limiting transmission. While we can debate whether lockdowns work better than other policies, the rigorous academic evidence is relatively clear — limiting activity limits transmission. But what this ignores is that no lockdowns were complete, and politicians appeared to have taken an almost arbitrary approach to the sectors they allowed to open at various times.

Counterfactually, what if they had decided instead to focus all of their efforts on the dual goals of limiting contact while keeping schools open? At least one possible answer for why they did not do this is because they did not do the political work necessary to generate demand for this policy among citizens.

What exactly did citizens want politicians to do? Along with colleagues, I conducted a study comparing the preferences of citizens and politicians for lockdowns (McAndrews et al. 2021). This study was conducted by the Media Ecosystem Observatory, among 2,527 and 1,502 Canadians in May 2020 and September 2020 respectively. We also partnered with the Canadian Municipal Barometer to conduct the same study among 665 municipal politicians in September and October 2020.[1]

Our study examined the tradeoffs citizens and politicians were willing to make between closing of different sectors and different levels of mortality from COVID. Complete details are available in our paper.[2]

In our study, we presented respondents with pairs of policy choices, which differed in how many and which of eight policies were enacted, and in how many deaths would be associated with that policy.

The policies included:

  • Tracking citizen movement with mobile phones;
  • Closing schools;
  • Closing non-essential businesses;
  • Closing non-essential government services;
  • Limiting public gatherings;
  • Requiring written permission to leave the home; and
  • The degree of government financial support to citizens.

The important findings are the following. First, citizens and politicians were both willing to forgo some policies — for example, written permission to leave the home — even if the trade-off was associated with a higher number of deaths. Second, citizens and politicians had very similar policy preferences. To a remarkable degree, they saw eye-to-eye on trade-offs. Third, however, there was one exception. Citizens and politicians were non-congruent on the policy of keeping schools open, with politicians less supportive of school closures than citizens. Importantly, as we show in a follow-up study, politicians systematically overestimated how much citizens supported keeping schools open.

By my lights, politicians’ policy preferences were the correct ones, and to the degree that citizens (and experts) disagreed, they were perhaps wrong. Keeping schools open, at least at the margin, was better for children’s mental and emotional well being and is surely associated with better learning outcomes. From the perspective of parents, and especially mothers, having children in school made working from home possible. It was thus also better for the economy. The inability of politicians to line up citizens’ preferences on this policy issue represents a fundamental policy breakdown on at least three levels. First, while politicians aligned with citizens’ preferences across a very large number of issues, they were not able to persuade them on arguably the most important one (or perhaps the second most important after limiting non-necessary gatherings between citizens). Second, because they overestimated support for keeping schools open among citizens, they likely did not engage in the persuasion necessary to make this policy the singular focus of lockdown policy. Third, they simply did not — in any province — make the rapid infrastructural changes, hiring or testing capacity ramp-up necessary to make in-person schooling possible. If they had made this a singular policy goal, perhaps they would have been able to take on this challenge.

In the early days of the pandemic, there was a widespread belief among policymakers that technology — specifically, the various tracking technologies embedded in mobile phones — could be used to help combat the pandemic. Broadly speaking, there were two possible ways such technology could be deployed. One is through monitoring the movement of citizens, and then using other controls to limit that movement. The second, substantially less invasive application, would be to use phones to alert individuals when they had been close to someone who was infected. The logic here is borrowed directly from more traditional forms of contact tracing. When Person A is infected, they make a list of everyone they had close contact with over the time period when they were infected and then asymptomatic, and those individuals are then contacted and encouraged to be tested. The limit to this is that individuals will not recall everywhere they have been or everyone they have been in contact with, and further still, it may be difficult to locate those individuals. However, if individuals’ phones can be used to retrace locations and times, and can be cross-referenced with other individuals’ phones, who are then automatically notified, then more efficient and effective contact tracing can occur.

Whatever the promise of contact-tracing apps, Canadian governments were both slow off the mark in developing apps and unsuccessful in generating widespread adoption. This represents perhaps their biggest missed opportunity.

In this case, we adopted a privacy-preserving act, which severely limited the amount of data collected on individuals’ movements, and made whatever data was collected essentially inaccessible to governments or public health agencies. Despite these privacy-first features, the COVID Alert app did not experience widespread adoption, severely limiting its usefulness, either in protecting individuals or in guiding public health strategies. Certainly, this lack of uptake meant that other restrictive measures could not be traded off in place of application adoption.

Could it have been otherwise? Why was there so little uptake? Some work in Canada (Rheault and Musulan 2020) and elsewhere (Horvath et al. 2020) suggests that privacy was a major concern for adoption of the COVID Alert application, and other similar applications. Another view, however, is that what hampered uptake was a lack of a different set of beliefs, first that government can use data and should use data to combat the pandemic, and second, that other individuals were going to use the app. Indeed, in a deeper analysis colleagues and I conducted on behalf of a project for the Schwartz Reisman Institute at the University of Toronto and the Center for Advanced Study in the Behavioral Sciences at Stanford University, we found just this (Loewen et al. 2021). Citizens’ beliefs about whether government should use data, and their beliefs that others would use also the app both mattered.

Using Media Ecosystem Observatory Data collected in the autumn of 2020, we first showed that individuals who had not yet adopted the app were significantly more likely to report a willingness to use the app in the future if they believed that government can positively use data to combat the coronavirus.[3] First, views about digital privacy generally did not matter for uptake of the app, either comparing those who had and had not downloaded the app, or prospectively among those who had not yet downloaded the app.[4] Moreover, political ideology did not consistently matter.

In addition to this belief increasing uptake on its own, it also increased uptake further among citizens who were told that adopting the app themselves could help others (including the elderly and those with underlying conditions).

Second, we showed that the belief that government can positively use data can encourage those who have not yet adopted the app to do so, conditional upon other people also using the app. To demonstrate this, we conducted an experiment in which we randomly varied information about how many people were estimated to have downloaded the app so far, and then remeasured the likelihood of respondents downloading the app in the future. Among those who do not generally believe that government can positively use data, the estimated likelihood of adoption does not increase as the hypothetical uptake rate increases. Instead, it remains steady at approximately 30 percent. However, among those who have a belief that government can positively use data, there is a steady increase as the estimated uptake rate increases. Indeed, among those with positive views about government data usage who are told the uptake rate is 10 percent, the estimated rate of adoption is 42 percent. But if those same individuals are told that the uptake rate is 50 percent, their likelihood of adoption reaches 50 percent. If they are told that the uptake rate is 90 percent, their own stated likelihood of adoption rises to 56 percent.

Clearly, the joint belief that government should use data and that others are taking up the COVID Alert app could have systematically increased uptake.

In retrospect, then, what could governments have done to increase uptake of the COVID Alert app, whether in its present configuration or in one that more effectively leveraged data? Our data suggest that two things needed to jointly occur. First, government needed to articulate how data could be used to better wrestle down the pandemic. In the specific case of this app, as configured, this was largely impossible. But government could still have made a more positive case for how it was using data to understand the pandemic, to target testing and to limit sectoral closings. Thankfully, it is never too late to increase citizens’ trust and imagination in how governments might improve their lives using data, but it does take an effort. Second, government could have affected a massive shift in citizens’ beliefs about how many other people were using the app. How could it have done so? One obvious measure would have been to offer to pay the cellular phone and data plan charges of all citizens who downloaded and used the app. Even the act of offering this service would have increased beliefs that others would use the app, effectively driving uptake among others. Given the massive costs of other income and business supports government was providing — largely as a result of forcing closures of businesses — there is likely a clear business case for their doing so.

The rate of vaccine uptake in Canada has been remarkable in the last two months. From May 1 to July 1, the percentage of Canadians with at least one dose of a vaccine has increased from 33 percent to 68 percent. By comparison, 44 percent of residents of the United States and 51 percent of residents of the United Kingdom had at least one dose. Those numbers have grown to 54 percent and 66 percent respectively. Canada, then, compares very favourably in vaccine roll out to some of its more well-known counterparts.

Such an analysis ignores three considerations, however. First, Canada’s initial roll out was very slow, attributable mostly to a lack of available vaccines. Indeed, as of Jan. 1, less than one percent of Canadians had received a vaccine. By March 1, well into Canada’s third wave, the rate was still under four percent. Had the government procured vaccines earlier, a large number of deaths would have been prevented.

Second, and more important, while Canada did do some limited targeting of vaccines, it was often both too narrow in who could be prioritized and then too broad in application. On the narrowness of targeting, many provinces initially prioritized health-care workers, but not teachers, and certainly not taxi drivers. On the overly broad targeting, often geographic targets were made at the level of forward sortation areas (the first three digits of a postal code). In urban centres especially, this can lead to wealthy neighbourhoods that are adjacent to poorer neighbourhoods being lumped into “hot spots.” Overly broad geographic targeting can prioritize those who do not need to be targeted.

Third, while Canada has played an important role in setting up and supporting an international vaccination sharing scheme (ie. COVAX), Canada has not yet contributed large numbers of vaccines to this effort. Indeed, we have nontrivial amounts of AstraZeneca vaccine currently sitting unused in pharmacy fridges.

Could Canada’s roll out have been different? There are two considerations here. The first is what was practically feasible, given the apparent organizational, data and logistical capacities of governments. But the second is what citizens would have supported. Would Canadians have supported different targeting in the roll out of vaccines? And would they have supported Canada taking a more active and generous role in providing vaccines internationally?

Two papers by colleagues and I suggest that the conditional answer to both is yes. To answer these questions, a team lead by Raymond Duch and Philip Clarke at Nuffield College Oxford surveyed citizens in 13 countries, on six continents, from countries representing more than half the world’s population. Our study was conducted in December 2020 and January 2021. Our goal in this study was to understand who citizens would prioritise for vaccine roll outs.

A first paper (Duch et al 2021) examined preferences for domestic vaccine prioritization. To understand this, we presented survey respondents with multiple pairs of individuals who varied across multiple characteristics, including their age, underlying medical conditions, income, ability to work from home and type of profession, with several different professions represented, including health-care workers, factory workers and taxi drivers.

What we found is that citizens in virtually every country prioritize the elderly and those with underlying conditions, but they also prioritize those working in service roles, such as taxi drivers, and those working in traditionally blue collar roles, such as factory workers.

Who do they send to the back of the line? Those with more money and/or those with the capacity to work from home. This finding was consistent across virtually every country. It was clear in Canada, and it stands in stark contrast to how much its own roll out has privileged those with the time, capacity and networks to find available vaccines. How different would Canada’s third wave have been with earlier, more broadly imagined, but narrowly administered, targeting? How much better would citizens have felt about themselves as a country?

A second shorter paper (Clarke et al 2020) examines preferences for global vaccine sharing. Here again, the results are crystal clear. Before the roll out of vaccines and whatever mix of euphoria and nationalism they have stoked, citizens in high-income countries were willing to see their countries share their vaccine supply with those in low-income countries. This was, again, a preference shared across most of the countries we sampled.

Perhaps the country lacked both the political bandwidth and appetite for bold domestic action and global leadership, but there was a different path to be taken by Canada on vaccine roll outs. The road it did take has been a remarkably successful one, especially in the last two months. But it was not the only one it could have walked.

“The political system is about change. It has to be about change. If you are not changing things you are of no value.” — Paul Keating, former prime minister of Australia

Canadians of all political backgrounds should be both proud and worried about the performance of their political and policy class during the COVID-19 pandemic. The purpose of this report has not been to celebrate Canada’s successes, but it must be noted that there have been successes. The performance of the Atlantic provinces has been exemplary, however much it is overlooked in central Canada. Across the country, public services have moved to transform service delivery and policymaking at a pace never before seen. And Canada has, in the aggregate, avoided the degree of bad outcomes seen in otherwise well-governed countries.

But there is still a cause for concern. On at least three scores — keeping schools open, using technology to combat the pandemic, and prioritizing vaccines — Canada’s political leaders have shown less political imagination than existed in their publics. What is the cause of this? It is likely some combination of complicated advice structures, a natural tendency towards risk aversion, and more lately, just plain fatigue. But some of the blame must also be on citizens, who could have demanded more of politicians, promised more reward for those who followed through, and held out forgiveness for those who failed in trying.

There will be lots of time for politics in the coming years, long after the pandemic is behind us. But there will be no shortage of mega problems to solve. Hopefully, one lesson Canada can learn from the pandemic is how governments can lead with bigger goals and bolder actions.

  1. We also ran the same study among provincial and federal members, and among Ontario school trustees. Our results do not differ substantively.
  2. The paper is currently under review and available by request.
  3. This was measured by creating an average response to three questions: “I am more concerned about government having my data than private companies” (reverse coded); “I am willing to give up some privacy in order to slow down the spread of the coronavirus”; and, “Government should use as much technology as necessary to slow the spread of coronavirus.” Response categories were strongly agree, somewhat agree, neither agree nor disagree, somewhat disagree, strongly disagree. The variable was split at its median, with those having above median agreement coded 1 and others 0.
  4. This is a measure taken as the average of the following five questions. “In general, how concerned about you about the following:” “Your privacy while you are using the internet”; “Online organizations not being who they claim they are”; “Online identity theft”; “People online not being who they say they are”; “That your credit card may be intercepted by someone else when you use it to buy something on the internet.” Response categories were not at all, not very much, somewhat concerned and very concerned.

Clarke, P.M., Roope, L.S., Loewen, P.J., Bonnefon, J.F., Melegaro, A., Friedman, J., Violato, M., Barnett, A. and Duch, R. 2021. Public opinion on global roll out of COVID-19 vaccines. Nature Medicine, 27(6), p. 935-936.

Duch, R., Roope, L.S., Violato, M., Becerra, M.F., Robinson, T., Bonnefon, J.F., Friedman, J., Loewen, P., Mamidi, P., Melegaro, A. and Blanco, M. 2021. Who should be first in line for the COVID-19 vaccine? Surveys in 13 countries of the publics preferences for prioritisation. medRxiv.

Loewen, P., J., McAndrews, J., Jad El Tal, K. K., Chen, H., Thompson, N., Moniuszko, A., Porco, E., Anderson, M. and Thompson-Bled, S. 2021. Why can governments use regulation, but not technology? Working paper.

Horvath, L., Banducci, S. and James, O. 2020. Citizens’ attitudes to contact tracing apps. Journal of Experimental Political Science. p. 1-27.

McAndrews, J., Lucas, J., Merkley, E., Rubenson, D., and Loewen P.J. 2021. Do Politicians and citizens agree on how to combat COVID-19? Evidence from Three Studies. Working paper.

Merkley, E. and Loewen, P.J. 2021. Anti-intellectualism and the mass public’s response to the COVID-19 pandemic. Nature Human Behaviour, 5(6), p. 706-715.

Merkley, E., Bridgman, A., Loewen, P.J., Owen, T., Ruths, D. and Zhilin, O., 2020. A rare moment of cross-partisan consensus: Elite and public response to the COVID-19 pandemic in Canada. Canadian Journal of Political Science/Revue canadienne de science politique, 53(2), p. 311-318.

Rheault, L. and Musulan, A., 2020. Explaining support for COVID-19 cell phone contact tracing. Canadian Journal of Political Science.