Advisory Council on Ethics and Modernization of Microdata Access - Meeting minutes for October 8, 2020

Date and location

8 October 2020, 1:00 p.m. to 4:30 p.m.
Virtual meeting via Zoom

Members

Anil Arora, Chief Statistician of Canada (Chair)
Cal Marcoux, Chief Information Security Officer, Canadian Institute for Health Information
Chantal Bernier, Privacy and Cybersecurity Lead, Dentons Canada LLP
David Robichaud, Ethics Professor, University of Ottawa
Len Garis, Researcher and Adjunct Professor
Mary Dawson, former Conflict of Interest and Ethics Commissioner, Lawyer
Robert Gordon, Executive Director, Canadian Cyber Threat Exchange

Guests

Dr. David Green, Professor, Vancouver School of Economics, University of British Columbia
Chantal Guay, Chief Executive Officer, Standards Council of Canada
Alain Beaudoin, Assistant Secretary to Cabinet, Results and Delivery, Privy Council Office

Statistics Canada guests/support

Jacques Fauteux
André Loranger
Greg Peterson
Lynn Barr-Telford
Gabrielle Beaudoin
Martin St-Yves
Eric Rancourt
Tom Dufour
Janique Godin
Geneviève Jourdain
Nicole Huard
Aiysha Abdullah
Farnaz Ahanin

1. Introductions and opening remarks

Anil Arora, Chief Statistician of Canada

Mr. Arora opened the meeting by welcoming members to the third bi-annual Advisory Council on Ethics and Modernization of Microdata Access (ACEMMA) meeting, and thanking them for their attendance. He noted that over the course of this year, increased prominence of outputs from Statistics Canada (StatCan) have doubled across all access metrics. He specified that these are direct results from increased partnerships and the work that StatCan is doing as data steward, such as creating a repository for personal protective equipment (PPE) inventory, the creation of multiple Dashboards, as well as performing contact tracing.

In addition to these achievements, of note were the implementation of strategic measures from the previous ACEMMA meeting in February 2020, including: refinements to the Trust Centre, modernizing access to microdata by considering public vs private institutions, and strengthening organizational agreements for access to health files. He outlined the importance of the Council's continued feedback, reminding them that the collective goal is to develop microdata access services and products that enhance the response to the needs of Canadians.

Mr. Arora invited Council members to continue providing their expertise and knowledgeable guidance on providing Canadians with timely access to data in new and innovative ways through an ethical approach where data remains safeguarded. Together, the Council will help the agency continue to achieve its mandate while respecting privacy, safeguarding information, and increasing collaboration through various modernization initiatives and new data sources in order to maximize impact.

2. Expert panel: Maximizing policy impact through data sharing and collaboration: the foundation

Dr. David Green, Professor, Vancouver School of Economics, University of British Columbia
Chantal Guay, Chief Executive Officer, Standards Council of Canada
Alain Beaudoin, Assistant Secretary to Cabinet, Results and Delivery, Privy Council Office
Gabrielle Beaudoin (Moderator)

Before launching into the Expert Panel discussion, Ms. Beaudoin invited each participant to introduce themselves, and share their relevant experiences related to Maximizing Policy Impact through Data Sharing and Collaboration. Panelists' presentations then began with Dr. David Green who shared a COVID-19 success story.

Presentation by David Green, University of British Columbia

Dr. Green shared his presentation on "Measuring the Viral Risk and Economic Impact of Sectors of the Canadian Economy" and expanded on his collaboration success story. He recounted to Council members that at the end of March, the Vancouver School of Economics were contacted by the British Columbia Centre for Disease Control. Dr. Green's team was recruited to help assess the economic impacts of the virus, examining the current state of the economy, risks the virus poses in particular occupations, rewards for keeping certain areas of business open, inequality effects, and joint modelling of the economy and the virus. He noted that Université du Québec à Montréal (UQAM) was also conducting similar research, leading the two teams to collaborate together on this work that served to inform decisions both in BC and in Quebec.

Dr. Green recounted how his team required data available in the Research Data Centres (RDCs) which were temporarily closed then as per public health directions to counter the pandemic. As a solution, Dr. Green's team collaborated with StatCan who provided immediate and substantial support for their research. Secure devices were shipped to the researchers (who were existing Deemed Employees), providing them secure access to the data required.

To ensure data safeguarding, new protocols were put in place in order for researchers to have the ability to collaborate with one another. Dr. Green's team utilized the data to build measures for rewards and risks for different sectors of the economy across Canada, with various characteristics measured. The data collected was then used to develop a risk scale using a factor analysis called the "Risk/Reward Assessment Tool". This tool is being used at both the federal and provincial levels. Dr. Green concluded his presentation in noting that the project was recently shared with StatCan to determine city-level information.

Presentation by Chantal Guay, Standards Council of Canada

Ms. Guay shared her knowledge and expertise on defining national standards. She touched on the universal nature of data and knowledge economy, and how the choices people make for data governance impacts health, well-being, and prosperity. The Standards Council of Canada (SCC) is the forefront voice of standards and accreditation on the national and international stage, playing a key role in innovation, and promoting safety while supporting businesses. She expanded on some of the challenges of standardization and the difficulty in regulating privacy, stating that perspective allows for a balanced approach.

Ms. Guay explained that SCC represents Canada at the International Organization for Standardization (ISO) and the International Electrotechnical Commission (IEC) connecting Canada to a variety of networks. Recently at the ISO, her team received approval to develop a management system under artificial intelligence (AI), for items such as linked data. They are also working to support a digital charter to benefit governments. She expanded that this falls under the realm of consent, privacy, cybersecurity, and governance. Ms. Guay touched on standards having the ability to bring different perspectives together, and explained that a roadmap is currently being developed to look at various standards and guidance, identify gaps that are essential to close, and the creation of a list of potential items to create. The roadmap is being developed by over 200 experts from academia, users of data, AI personnel, and ethics experts, in order to broaden the representation of perspectives.

Ms. Guay continued to touch on how data governance are abstract concepts, however, they play an important role in the collection, sharing, and use of data. For these abstract concepts, Ms. Guay and her team have decided to look at 3 use cases: community health with StatCan using Covid-19 testing and how results are reported, digital identity and open banking where transactions are restrictive, and consumer food safety in collaboration with the Canadian Food Inspection Agency. Ms. Guay concluded her presentation in quoting five recommendations on developing the standard setting process: short and simple, bigger is not always better, transparency works, delivery matters, and digital leadership, technology, and money are not the solution but rather people and digital leaders are.

Presentation by Alain Beaudoin, Privy Council Office

Mr. Beaudoin shared his knowledge of Privy Council Office (PCO) initiatives, specifically the result and delivery unit, which was created to support the Prime Minister on government implementation, and to encourage the use of evidence to inform policies and programs on an ongoing basis.

He expanded on PCO's stance on encouraging data collection, but more importantly, collecting the right data and the key data points. PCO works closely with departments to increase collaboration, but also to increase transparency and to improve data sharing for various types of data. As an example of these efforts, Mr. Beaudoin touched on two instances where Covid-19 highlighted the importance of data collection: firstly, the ability to provide services in a digital way to Canadians and businesses, and secondly, the impact of Covid-19 compared to other previous recessions. What the data showed was that the virus had an uneven impact on vulnerable populations, women in the labour force, specific occupations, and regions. Mr. Beaudoin elaborated that this pandemic increased the focus on data, the story behind the data, and how to measure the success of programs.

Mr. Beaudoin brought forward a question for Council members: how do we create a shift in culture that is not just reactive, but will lead to a fundamental change of how to look at microdata as a solution moving forward? As we move through this pandemic and aim to continue economic recovery, Mr. Beaudoin emphasized the importance of access to microdata, and ensuring resources are invested where they will make a difference.

Mr. Arora closed the introduction period stating "data, unlike oil, is something that increases in value as it is shared." There has been a paradigm shift; it is about how we share data, and how to do it responsibly to ensure we protect privacy and confidentiality, maintain standards, and communicate value proposition.

Panel discussion: Q&A discussion period

Gabrielle Beaudoin (moderator)

Question 1: We talk about a secure and successful data sharing model. What would improve partnership/collaboration, especially in this COVID-19 environment?

Dr. Green responded that the collaboration efforts that were done with his group at UBC worked very well, and that it fully protected confidentiality, and also gave the researchers access more rapidly and in a way in which they could work with the data around the clock while the Research Data Centres (RDCs) were closed due to Covid-19. He noted that the Statistics Act is clear and strong enough to ensure researchers apply the approved protocols. Dr. Green indicated that he sees the potential of the Cloud for the foreseeable future, and that we should question why this isn't the current reality of conducting research.

Council members responded by adding that lessons learned would be helpful and a foundational piece to a standards model going forward, a model which could be duplicated.

Question 2: What challenges do you foresee in regards to data sharing and collaboration across jurisdictions? How could those challenges be overcome? How do we gain public confidence with respect to data?

Mr. Beaudoin began by expanding on how there is a desire to share outcomes to increase confidence with the public, and showcase what the government has done and how they have progressed. He questioned how we can figure out a way to make data open to the public, and empower individuals and researchers, while benefiting from the input. Mr. Beaudoin added that it is important to work with departments and agencies to make the data available to the public, and that this allows everyone to benefit from the output. He emphasized that this is the dialogue that we need to continue holding with experts to improve our policies and programs.

Council members raised that if standards are not judiciously implemented, it can accumulate in the form of additional costs, such as duplication of reports due to varying certification processes across provinces. It was raised that the potential power of data can be hindered when there are a lack of standards and regulations; as an example, researchers who require data from both the provincial and federal levels may face legal barriers between jurisdictions, and data barriers between collection methods. As such, members stated that data privacy rules and oversight models must be aligned as much as possible. Ms. Guay specified that SCC's goal is to align with federal, provincial, and territorial governments.

Concerning public confidence, Council members stated that there was an urgency to build upon commonalities for data. With the pandemic, there has been a sense of urgency to build partnerships and collect data in the interest of the public. Building upon the commonalities of evidence based decision making, and shared principles, would allow the Government to use data when it is in the interest of the public. There are calculated risks to being agile, but data driven decisions should be an aspect that is focused on. The Council advised that these data success stories, such as COVID-19, should be shared publicly to encourage public confidence and build an understanding of how data can drive policy and create stability. Data should be readily available to policy makers, and those managing health.

Mr. Arora concluded the Panel Discussion question period by stating that data is a team sport. Mr. Arora emphasized that having a data trust model allows data to be collected once with transparent assurances, and to create the linkage environment to bring forward the cross-sectional insights and highlight the quality. He continued stating that the notion of data trust and stewardship adds certainty and responsibility to the trust of Canadians when sharing data. Mr. Arora finished his statement by expanding on how our policy makers and decision makers need the data, and that Statistics Canada needs to categorize, understand utility, and check technology and monitoring, in order to assure Canadians that we are compliant to our promises and that their confidentiality and privacy are protected.

3. New data sources and methods, and inherent ethical questions

Eric Rancourt, Director General, Modern Statistical Methods and Data Science

Mr. Rancourt discussed data sources, data stewardship and trust roles, ethical questions, and ethics principles being examined in the process of data collection and management at StatCan. He explained to Council members that with society being more complex, StatCan wants to equip Canadians by informing them fairly, transparently, and allowing for autonomy. The Council acknowledged the importance of communication with Canadians in order to build trust, and the importance of using trusted data source to make informed decisions. Mr. Rancourt emphasized that StatCan works for the public good, and also a moral duty to find the truth, unbiased information, and quality statistics, while maintaining a good representation of Canadians through a proportionality-based approach. Council members advised that transparency surrounding data collection, and publicizing results accessibly are two key steps in meeting public interest.

Mr. Rancourt expanded on how organizations have risen to the challenge of standards by reviewing needs from an impartial, ethical, and transparent view. He added that while StatCan has the Necessity and Proportionality Framework containing an ethical component, the agency is trying to instil ethical thinking from the beginning to produce ethics by design. While Mr. Rancourt identified that there was a need to explore new areas of data and data collection, there are a number of values that must be considered when attempting to engage and inform the population, such as: autonomy, consent, liberty, and perceived desired level of privacy.

Discussed were new methods of data collection including web scraping, and mobility data. Regarding the use cellphone data, Council members stressed the importance of Privacy Impact Assessments, as it would push StatCan to question what is truly necessary and assess whether cellphone data use is proportionate to public interest. Mr. Rancourt explains that there are many companies who can provide this information; however, this brings forward a distinct issue of fairness and representation. Alternatively, Council suggested that StatCan could consider turning to social media and using machine learning to analyze and asses the data on various platforms, in accordance with the Privacy Act.

Mr. Rancourt concluded his presentation by leaving Council members with the following question: How do we consider ethics and still provide unbiased information?

Council agreed that as a society, we have both an ethical obligation to ensure we are not overburdening or over representing at the expense of others, and that we do not end up with an absence and lack of data needed. Overall, Council members would like to see that StatCan become explicitly mentioned in the discussions surrounding data. Council members agree that the moment is right to teach the public what we are doing with the microdata, and to help them understand what it means. It is important to bring Canadians the value of StatCan's work, which may encourage a change in rapport. Council provided Mr. Rancourt with final advice to keep in mind: privacy, transparency, and fairness are key.

5. Closing remarks