Canada’s battle against the Covid-19 virus has reached the stage where it is now possible to consider how successful our fight has been and at what cost. Reflecting on past efforts provides a better understanding of what has been accomplished so that we can deal with future pandemics in the most efficient and effective manner. To do this in a rigorous and scientific manner, we will rely on the familiar economic framework of cost/benefit analysis. Cost/benefit is widely used across the public and private sectors as the way to select the most cost-efficient option among several possible courses of action.
While some might say that cost should be no object when it comes to protecting people from disease or saving lives, this is not necessarily so. If we adopt an excessively costly or wasteful approach to fighting one battle, it deprives governments and other organizations of resources required to pursue other worthy objectives. Doing things in the most cost-effective manner possible is therefore always desirable, and should be considered the mark of good public policy.
Let us begin with the costs. Many of the costs arising from the coronavirus pandemic involve a direct personal impact on individual Canadians. They are the result of government policies requiring residents to self-isolate, keep a social distance from others and limit public activities. Government orders have also closed numerous businesses, interrupted the provision of hospitality, travel, and education, postponed elective medical procedures in hospitals and prohibited visits by families and friends of loved ones in care homes for the elderly.
Obeying these requirements has caused widespread unemployment. It has also dramatically increased the incidence of loneliness, depression, interpersonal conflicts, divorce, alcohol consumption, pain from untreated medical conditions, substance abuse (including fatal drug overdoses) and suicides. While this constitutes a very important psychological and physical burden on all individuals, it is impossible to properly put a concrete value on these issues. Under a strict cost/benefit framework, we can only offer token, if heart-felt, recognition of the hardships involved.
Government lockdown policies, however, have also led to significant economic costs which can be measured in a more precise manner. In particular, there is the loss in output, or gross domestic product (GDP), suffered by Canada’s economy. This figure reflects all the restaurant meals not produced, construction not performed, educational services not delivered and so on.
The loss from this foregone production will never be recovered; it is gone forever. The magnitude of the loss can be found in forecasts of GDP. For our purposes we will use figures from the International Monetary Fund (IMF). According to its most recent estimates, Canada’s GDP in 2020 will be 6.5 percent below that of 2019. Since GDP in 2019 GDP was $1.7 trillion, the value of reduced output in 2020 is thus $113 billion.
The cumulative economic costs arising from the coronavirus will undoubtedly be much higher than this as any recovery is likely to take much longer than one year. The OECD, for example, recently issued an alert warning of a possible second wave of Covid-19. Given expectations for a multi-year global economic crisis, it seems reasonable to assume the total cost for Canada will be at least $226 billion, or twice what the IMF calculates as the single-year hit.
It is important to note that government payments to individuals or businesses to prevent financial suffering and bankruptcies – including such things as the CERB, corporate subsidies and many other new government programs – are not considered a cost in this analysis. Despite the massive deficits presently being incurred, these extra public expenditures do not necessarily lead to a loss of output over the short term. Rather, these payments add to the public debt, which in the future will require interest payments made by taking money from some Canadians (taxpayers) and transferring it to other Canadians (those who hold the debt).
Having established a cost estimate at $226 billion, let us now turn our attention to estimates of the benefits delivered by the government’s pandemic-fighting policies. This could include reductions in the number of infections as well as avoided deaths. Infections cause symptoms that range from non-noticeable to sufficiently serious to require hospitalization and, in some instances, eventually leading to death. However since it is difficult, if not impossible, to measure the value of suffering avoided by government policies in any comprehensive way, non-fatal infections will be ignored in this calculation.
It is far simpler and clearer to concentrate on deaths avoided as the true benefit of government policy. Computer models produced by infectious disease specialists predict the number of deaths likely to occur with and without mitigating government policies, the difference representing the number of deaths prevented. Here we will rely on the widely-cited Imperial College of London (ICL) model. While controversial for some, most criticism of this study has focused on its tendency to exaggerate the potential death toll; as such we feel confident using it as a base case since it is likely over-estimating the benefits.
In the absence of an effective vaccine, the ICL model predicts four deaths per thousand population if “non-pharmaceutical interventions” (social distancing, quarantining, etc.) are implemented, and eight deaths per thousand without these inventions. In other words, the ICL model predicts that mitigation will half the number of fatalities. Given that the actual number of deaths in Canada as of the middle of June was 8,300, this means the estimated number of deaths prevented is also 8,300.
This figure, however, does not take account of the fact that pandemic-fighting policies have increased unemployment, which is widely acknowledged to increase the prevalence of suicides. A recent study projects Canada will experience approximately an additional 2,100 suicides due to the coronavirus. Subtracting these deaths from the 8,300 fatalities prevented by government policies yields 6,200 net deaths prevented. Not considered in this calculation is the fact more than 90 percent of the deaths in Canada attributed to the Covid-19 virus involve individuals over the age of 60, many of whom already had shortened life expectancies due to pre-existing medical conditions.
Costs Outweigh Benefits Four-to-One
Having separately estimated the costs and benefits of the government response to the coronavirus pandemic, we must now bring our two calculations together. Dividing the total estimated cost of $226 billion in lost output by the 6,200 lives saved produces an estimated economic cost of $36.4 million per life saved. Again, this estimate is based on a series of assumptions enumerated above, all of which can be readily challenged. But any reasonable tweaks to costs incurred or lives saved seem unlikely to change the basic conclusion. So is an expenditure of $36.4 million for each life saved to date from Covid-19 a reasonable or cost-effective policy?
Economists and statisticians have long used sophisticated models to calculate the value of life for a wide range of public policy purposes. It is impossible to properly assess road safety initiatives such as lower speed limits or better lighting at intersections, for example, without applying a dollar value to the lives expected to be saved by such interventions. The underlying concept here is called the value of statistical life, a calculation that attempts to put a dollar figure on the value of lives saved through government policy. In Canada, the value of a statistical life prescribed for federal cost/benefit analysis was set at $6.11 million in 2004; after adjusting for inflation this is about $8 million today. (Interestingly, the lifetime earnings of the average Canadian were $2.7 million in 2017.) This means that any intervention costing more than $8 million per life saved would fail Ottawa’s standard cost/benefit requirements.
Some readers will no doubt take issue with the use of such a dispassionate method of assessing government policies. For the bereaved, no amount of money can ever properly reflect the loss of a loved one. And yet these calculations are made in the spirit of similar estimates used by government agencies across Canada and throughout the world on a regular basis. It is a way of estimating whether we are applying resources efficiently towards our goal of saving lives. If it turns out that we are not, it suggests that we could have saved more lives by doing things differently. Surely we can all agree that this would be a worthy goal?
Setting aside any understandable sentimentality about measuring life using purely statistical methods, the results described above reveal the cost per life saved due to the current suite of coronavirus precautions of $36.4 million far exceeds the $8 million benchmark set by the federal government’s own regulations. Costs incurred are thus more than four times larger than would be acceptable if the fight against the coronavirus were any other government initiative. With respect to the robustness of this result, note that even if we did not double the IMF’s 2020 prediction for the economic loss experience by Canada to cover multi-year implications, our calculation of the cost per life saved ($18.2 million) would still be more than twice the $8 million benchmark.
It is obviously too late to use this information to influence lockdown policies already implemented and now being gradually lifted. But best practices demand that these results be incorporated into plans for fighting similar crises in future. So how can we ensure a more efficient response to the next pandemic – including the possibility of a second wave of the coronavirus – that is in line with existing federal cost/benefit procedures? One possibility is to consider the approach adopted by Sweden.
A More Efficient Approach
Sweden is widely recognized as having met the challenge of the global pandemic in ways that differ markedly from most of the rest of the world. Most importantly, it has avoided a general shutdown of its economy. Its policies have focused on reducing the number of deaths rather than of infections. This is an approach that accepts three basic premises, all of which appear to hold true in both countries.
First, deaths caused by the Covid-19 virus are overwhelmingly concentrated among the elderly. According to the most recent government statistics, 72 percent of all Canadians who have died from the coronavirus were over the age of 80; a further 25 percent were aged 60 to 79. The vast majority – 97 percent – of Canadians who died from the virus were thus older than 60.
Second, the overwhelming majority of the infected population experiences symptoms that are mild, do not require hospitalization or lead to death. The validity of this premise follows from the fact that 97 percent of all deaths are among Canadians older than 60. Among younger patients, the hospitalization and mortality rate is remarkably similar to those experienced every year due to seasonal influenza. The fact that the death toll from seasonal flu has never been considered serious enough to warrant costly mitigating policies beyond public service messaging recommending vaccination can be considered supporting evidence in this regard.
Third, individuals who have recovered from the infection are immune, will not infect others or become ill again for some significant length of time. This premise is valid if the Covid-19 virus has much the same effects on immunity as other viruses that have plagued humanity in the past. It is highly likely that this premise is correct, but more research is needed to verify it.
Assuming these premises are valid or require only modest modification, the Swedish model appears to present an effective and less costly alternative to the policies adopted in Canada and most of the rest of the world in the first half of 2020. This includes encouraging people to work from home and to practice good personal hygiene, keep social distance, self-isolate when ill and not attend gatherings involving more than 50 people. Strong and extensive measures to protect the elderly from contact with infected individuals are also recommended.
Significantly, the Swedish model does not require that extensive areas of economic activity cease. Factories, restaurants, gyms, educational institutions, public transit, construction, and other businesses remained open throughout the crisis. This decision limited the damage to output and will reduce the overall cost. If adopted in Canada, even the most expensive procedures to protect the elderly would come to a small fraction of the estimated total of $226 billion incurred by our nation. The same goes for any additional cost for treating younger patients in need of hospitalization due to a higher infection rate in the general population.
Sweden offers real-world evidence of the efficacy of such an approach. The full evidence on the economic savings will become available only over time, but the most recent available data indicate that the loss of output was far less severe in Sweden than Canada. In the first quarter of 2020 Sweden’s economy grew by 0.1 percent while that of Canada fell by 2.1 percent. Other factors may have contributed to this difference, such as the decline in Canada’s energy industry, but are unlikely to be significant enough to negate the validity of the comparison. So the costs borne by Sweden are substantially lower.
On the other hand, benefits are also lower. Current figures for deaths per 100,000 due to Covid-19 stand at 49.6 in Sweden and 22.6 in Canada. While Sweden’s figure is below that of the hardest-hit European countries including Britain, Italy and Spain, the large difference with Canada is puzzling because it seems likely the age-pattern of deaths is the same in both countries. If so, this implies a difference in procedures regarding protection of the elderly between Sweden and Canada. Why this may be the case raises important questions for future research. Are there differences between the two countries in how they count Covid deaths, the proportion of the elderly in the population, the design and staffing of care homes, the resources devoted to insulating the elderly and health-care providers, and so on?
Whatever the flaws of the Swedish approach, however, we cannot ignore the fact Canada’s experience explicitly fails the requirements of sound, cost-effective public policy as demonstrated by our simple cost/benefit analysis. With calls from many quarters that our approach to this and future pandemics must be guided by “science” – and given the grave implications for the economy arising from policy mistakes in this area − we should be mindful of the need to adopt an analytical approach using the most relevant and effective economic tools.
Herbert Grubel is Emeritus Professor of Economics at Simon Fraser University and a senior fellow at the Fraser Institute.