Rural Futures: Data & Decisions

By: William Ashton
Published: May 7th, 2020

Data informed decision-making is not new to government or public policy. What data and how it is interpreted is critical to strategy building and actions. The emphasis on a go-slow approach in Canada is a strategy with built in assumptions about the role of citizens. But there are other strategies.

Evidence-based decision-making has a strong tradition in rural policy-making, and public policy in general. In Canada, we have seen decisions by the Liberals guided by evidence, which is in contrast to an earlier Conservative Government that had Federal scientists demonstrating against the Government’s reduction in their numbers and blatant disregard for their work in informing public policy (Pedwell, 2012). Informing decisionstrategy in Canada s with analysis is common in the medical profession, where evidence consists of quantitative data from scientific and mathematical calculation that supports gauging the strength of inductive arguments for guiding decisions. 

Evidence-based practice is top of mind today in the face of COVID-19 as experts develop the models that governments depend on to guide the right resources to the right places at the right time. Such an approach seems clear and uncomplicated, but is it? In Canada, early in April 2020, citizens injecting themselves into this crisis and asked to scrutinize such evidence government was using, while others tried to piece together a path forward for government actions by examining other jurisdictions. 

Recently in Canada, public debate has focused on models forecasting the pandemic’s potential spread. This serves as a case in point for a model’s framework of data, decisions, and dominos. Any forecast includes assumptions, and the best assumptions are based on data from past trends. SARS, other infectious diseases, and the 2008 financial crisis swept across Canada, yet we are told that COVID-19 is different. Public health advisors want to ensure sufficient equipment is available for the ‘hot spots’ where outbreaks are the worst. A model helps identify possible locations and the number of cases. The hold-back is often ensuring that assumptions composing the model are reasonable, so in turn the forecast reflects some degree of reality. 

As with many things in life, even though we are talking life and death here, models based on weak data sets and unlikely assumptions can at times seem more like a guess, than a prediction. As assumptions and data sets improve and become more robust, the forecast results formulate into future situations or labeled as scenarios. Walsh and Hunter (2020) report Canada lagging behind in early April in releasing a national forecast based on multiple data points over time. Ideally we want to understand changes in four critical measurements of the ‘COVID-19 box’: with more testing comes more accuracy, while keeping citizens isolated in their own homes, initiating rapid tracking of those related to each confirmed case, and ensuring that travellers entering the country are quarantined. Meanwhile, other governments around the world may be relying on incomplete case data of those surviving and falling victim to COVID-19, which they know is incomplete but still must use to make and update their evolving forecasts as data continue to be collected and shared. Governments are seeking to understand the spread of the pandemic in order to guide decisions – health decisions, equipment decisions, and economic decisions. 

Even given incomplete data, the Canadian Government’s forecast was not the only one in Canada. With Provinces sharing responsibility for health services with the Federal government and citizens, Provinces generated their own models, assumptions, and case data by early April 2020. The pressure was building to release forecasts by both levels of government, as countries such as New Zealand and USA released their planning scenarios, often focusing on the blunt impact of this disease in terms of death counts. As reported on April 3, 2020, what seems important in Canada is having some consistency across jurisdictions, which required coordination and cooperation among health specialists and modelling teams, coupled with data sharing (Walsh and Hunter, 2020). As the first week of April closed, Canadians were “still in the dark about what information government is basing its plans on”, report Walsh and Hunter (2020). By April 10th at the provincial level, British Columbia released its modelling, followed by Ontario, with subsequent releases by New Brunswick and Alberta. 

Besides the multi-jurisdictional aspects of such modelling by governments, citizens are part of this discussion too. Experts say that releasing the models, knowing they are imperfect, nonetheless builds trust with citizens and helps governments get buy-in for unprecedented and restrictive policies, including self-isolation (Walsh and Hunter, 2020). The release of the forecasts are as much about making a statement that governments and the front-line health workers are not fighting this virus blindly or alone. 

With schools and businesses closed, and an entire way of life disrupted by travel bans and physical distancing, citizens are asking governments to be more transparent. The former health minster Jane Philpott calls for “radical transparency.” (Picard, 2020) Certainly, models bring the stark reality out in the open. What may be easily overlooked in the calamity of this crisis is that citizens are, in all probability, already discussing worst-case situations as many are laid off, and others are losing their life savings on their shuttered businesses, even as a few are rapidly hiring new workers with exploding sales. 

What citizens lack are specifics provided by a forecast. Some leaders see the value of an informed constituency, even in such turbulent conditions, and even if the modelling is a real wake-up call for us all. People understand this is a serious and dynamic situation, which is why they need to be fully informed as they do their share to reduce the spread of the virus. By April 10, 2020, upon reviewing recently released federal and provincial forecasts, Globe and Mail reporter Picard had two main observations: citizens are making and will continue to make the difference between the best and worst scenarios, depending on their adherence to health guidelines; and second, the projections had little in common.

Different death rates and totals, and some not reporting death rates (like B.C.), and their focus on capacity of health services including intensive care. Quebec did not include a longer-term projection to the end of summer, stating, “forecast beyond April 30, 2020, is tantamount to practising astrology.” If citizens are important in the forecasts, the reports make no mention of the assumption about what percentage of citizens have to be practicing lockdown. From earlier polling, Picard (2020) noted that one in four or 25% of Canadians were not practicing distancing. It begs a question about what level of compliance with public policy is needed to bring about success. What percentage of citizens need to cooperate to defeat COVID-19: 75%, 90%, 100%? And comply for how long? While forecasts are not crystal balls exhibiting the future, they help imagine a near future and what could happen; and thanks to the attempt to do so, as one citizen, I feel more informed and willing to comply than I would if I did not have a forecast. 

If COVID-19 were a case study for a graduate class in policy making, data, and imagining futures, this Note could be used to outline the challenges and nuances among stakeholders and activities over time, coupled with the role of expert advisors, and uncertainty. Even with evidence, a forecast in itself is not a decision or policy. Clearly there is not one decision required in these forecasts, but many decisions by many, and it is an ever-changing situation in Canada and elsewhere. The stakeholders encompass both levels of government (federal and provincial), since both have responsibilities for health services in Canada. Too often the largest group of stakeholders – citizens – is left out, but not in this case. Indeed, the spread of this disease hinges on what citizens do or not do, in the first wave and subsequent waves. 

As governments seek more data to shed light on the situation, their current and evolving forecasts carry uncertainty and risk, as I’ve remarked. Here uncertainty includes not knowing in absolute terms the trajectory of infection rates and death rates, for example, while at the same time being hampered by a lack of testing to improve accuracy in reported infection rates. Often when making public-policy decisions, a gap exists between what the data are explaining and what is a reasonable response. 

Policy analysts help to distinguish among three situations: the easiest is where a clear decision needs to be taken, e.g. how many surgical masks need to be shipped to a hot spot? They also differentiate a situations that is a problem, meaning data are clear and the solution can be defined. The third situation is the hardest: a dilemma with multiple situations that are wicked or complicated and often interrelated, requiring multiple decisions over time, so there is no single solution or policy. Rather there are interactions over time, and all the while decision-makers must hope they do not make the situation worse (Lowy, 2007). 

Without accurate data and trends on infection rate, a single projection line with high confidence is simply not possible with COVID-19; thus there are many assumptions and unknowns. Managing the associated risk is equally important, and there is no bigger risk than the human factor – what will citizens do? Based on governments asking and demanding compliance, there are still too many citizens taking risks by being in groups and not physically distancing. These few are putting all of us, ‘the commons’, in jeopardy. The provinces responded with new laws and fines hoping to modify behaviours of the few non-conforming citizens, be they in rural or urban areas. 

In terms of the scale on which a model is based, such as the provincial or national scale, the above discussion addresses both. The respective governments seem, in general, to share a similar logic, meaning with the right amount and type of data, a strong forecast is possible. Meanwhile, Waldie (2020) reports on a different kind of model, the one being used by Prime Minister Lofven of Sweden. The PM’s overarching assumption is “keeping the country largely functioning and aiming health measures at the most vulnerable”. Knowing COVID-19 has staying power for 12, 24, and more months, Sweden is taking an approach to slow the economy, while maintaining the potential to start reinvigorating it very quickly once the crisis is over. With volunteer measures of travelling less, working from home, and social distancing, along with preparation measures in elderly homes and hospitals, Sweden’s go-slow approach is in sharp contrast to their Scandinavian and European counterparts. Yet it fully recognizes that citizens make a significant difference. Waldie (2020) cites a survey that shows Swedes trust the government and each other to a degree rarely seen in other countries, with more than 60% of the population in agreement. Equally important, Waldie (2020) states, “the government and institutions generally speaking trust citizens to do the right thing.” Another contributing factor in this approach is that almost half of Swedish households consist of a single person, where 30% is common across Europe. Context and culture are heavily factored into Sweden’s approach. Their model of go-slow seems to suggest a significant weighting and confidence placed on the role of cooperative citizens, as compared to other countries.

References Cited:
Lowy, A. (2007). No problem. Bloomington, IN: AuthorHouse.

Pedwell, T. (2012). Scientists take aim at Harper’s cuts with ‘death of evidence’ protest on Parliament Hill. Global and Mail, July 10, 2012. See: https://www.theglobeandmail.com/news/politics/scientists-take-aim-at-harper-cuts-with-death-of-evidence-protest-on-parliament-hill/article4403233/

Picard, A. (2020). We’ll all just have to accept that even the best forecasts will be imperfect. Global and Mail, A7. April 10, 2020.

Waldie, P. (2020). Sweden has faith in its pandemic response. Europe? Not so much. Globe and Mail, A7. April 3, 2020.Walsh, M., and Hunter, J. (2020). Ottawa holds back coronavirus projections. Globe and Mail, A3. April 3, 2020.