Statistics Canada Data Strategy

• A message from the Chief Statistician of Canada
• Summary
• Activities at a glance
• Statistics Canada Data Strategy Framework
• Drivers, challenges and goals
• Data governance
  • Short-term activities
  • Medium-term activities
  • Future activities
  • Expected outcomes
• Data stewardship
  • Data discovery
    • Short-term activities
    • Medium-term activities
    • Future activities
    • Expected outcomes
  • Data digitalization
    • Short-term activities
    • Medium-term activities
    • Future activities
    • Expected outcomes
  • Data interoperability
    • Short-term activities
    • Medium-term activities
    • Future activities
    • Expected outcomes
  • Data management
    • Short-term activities
    • Medium-term activities
    • Future activities
    • Expected outcomes
  • Data resources
    • Short-term activities
    • Medium-term activities
    • Future activities
    • Expected outcomes
  • Data trust framework
    • Short-term activities
    • Medium-term activities
    • Future activities
    • Expected outcomes
  • Data leadership
    • Short-term activities
    • Medium-term activities
    • Future activities
    • Expected outcomes
• Next steps
• Glossary of terms
  • Data ecosystem
  • National statistical system
• References

A message from the Chief Statistician of Canada

In 2018, Statistics Canada marked its centennial: 100 years of service in managing confidential data from Canadians and using its expertise to transform these data into meaningful insights on Canada's economy, society and environment. The agency's activities have always been driven by sound and robust data strategies to address emerging needs while maintaining the trust and confidentiality of Canadians.

Statistics Canada has a mandate to promote and develop integrated social, economic and environmental statistics pertaining to Canadians, the provinces and the territories. This mandate includes collaborating with government departments and agencies to collect, compile and publish statistical information, including information derived from departmental activities; to promote integration; and to avoid duplication of the information collected. The agency provides unbiased, high-quality data that are integral to the well-functioning of society and a statistical framework upon which the nation makes decisions.

In Canada, and globally, data and the digital economy are driving innovation, along with economic and social opportunity. Businesses, governments and citizens are seeing an explosion in the volume of data and data analytics technologies. But more data does not automatically result in greater knowledge and increased prosperity and social inclusion in Canada. This new data environment poses real challenges to governments and citizens. For example, technologies such as artificial intelligence (AI) and machine learning (ML) are reshaping analysis and have the potential to change the basis of decision making. The advent of "alternative facts" from dubious sources and a growing disregard for evidence and analysis in political and civil discourse all emphasize the need for greater data literacy and access to objective evidence for all Canadians. Statistics Canada recognizes that as institutions evolve in the midst of a data revolution, and as citizens are bombarded with large volumes of data from a variety of sources, the need for trusted, high quality data and competent data stewardship has never been greater.

Capitalizing on the data revolution will require a number of key enablers, including a modern, adaptive and agile statistical agency. Statistics Canada's data strategy is embedded in its modernization agenda. The agency is actively researching, developing, and implementing solutions to provide Canadians and decision-makers with increased access to high-quality, relevant and timely information on issues affecting our society using methods that preserve privacy, confidentiality and security; to provide expertise to partners, users and citizens; to provide trusted data insights and standards; to decrease response burden and capitalize on existing administrative data (non-survey based); to use innovative data collection approaches, such as crowdsourcing; and to expand and leverage innovative partnerships. Statistics Canada must continue to work closely with partners and stakeholders to remove barriers to data sharing, including technical, cultural and legislative.

Over the past year, Statistics Canada's modernization included investments in data infrastructure (for the development of the Data Analytics as a Service platform), improved gender-based analysis, and statistical capacity building for Indigenous peoples.

Statistics Canada's ongoing modernization and expertise as a world-leading data steward will provide great value to Canadians. In particular, Statistics Canada is committed to:

• improving transparency and openness;
• continuing the development of innovative methods to preserve privacy, confidentiality and security;
• increasing access to data and microdata to create information that drives innovation, inclusion and drives insights;
• increasing data literacy and statistical capacity building; and
• leading and fostering the strengthening of the national statistical system.

The Statistics Canada Data Strategy (SCDS) will be an evergreen document to provide a course of action for managing and leveraging the agency's data assets to ensure their optimal use and value while maintaining public trust. As the nation's trusted provider of high-quality data and information to support evidence-based policy and decision-making, the SCDS also naturally extends into the agency's course of action for supporting and providing expertise on data to other government organizations (federal, provincial, and territorial), non-government organizations, the private sector, academia, and other national and international communities. Statistics Canada coordinates the national statistical system (NSS) working with provinces, territories, municipalities, the private sector and non-government organizations Statistics Canada seeks the expertise and counsel of many advisory groups and has a comprehensive and well-established system of committees to ensure oversight and robustness of the system

This document provides a roadmap for how Statistics Canada will continue to govern and manage its valuable data assets as part of its modernization agenda and in alignment with and response to other federal government strategies and initiatives including the Data Strategy for the Federal Public Service, Canada's 2018-2020 National Action Plan on Open Government, and the Treasury Board Secretariat Digital Operations Strategic Plan: 2018-2022.

Finally, I would like to thank Statistics Canada employees for drawing on their collective expertise to guide the creation of this strategy. The SCDS will ensure that core functions of the agency, in the form of strategic data capabilities, will continue to make data more accessible, drive innovation, and ensure evidence-based decisions in support of an inclusive and data-driven economy and society while maintaining public trust.

Summary

Over the last 100 years Statistics Canada has evolved from a centralized statistical agency producing data tables on the state of Canadian economy and society to a modern, dynamic organization producing insights which result in data-driven policy decisions and better services for Canadians. As Canada's national statistical office, Statistics Canada has a clear vision: to deliver insight through data for a better Canada. The agency's role in a whole-of-government approach to data stewardship and data management is to ensure the availability of high-quality, trusted data to inform government programs and services, while safeguarding information and protecting the privacy of Canadians. Further, it is also the agency's role to help public servants and Canadians become more data savvy. Canadians need high-quality, insightful, accessible and timely data. The agency is committed to leading the integration of statistics and strengthening the national statistical system by enabling the strategic use of data and delivering data insights through innovative collaborations and partnerships.

Statistics Canada is achieving its vision through its modernization effort, which was initiated in 2017. Modernizing Statistics Canada's workplace culture and its approach to collecting and producing statistics is resulting in better and faster access for Canadians to the statistical products they need. Specifically, the goals of modernization are to:

• produce more timely and responsive statistics—ensure Canadians have the data they need when they need them;
• develop and release more granular statistics to ensure Canadians have the detailed information they need to make the best possible decisions;
• raise awareness about Statistics Canada's data and provide seamless secure access;
• provide leadership and expertise in strengthening the national statistical system—improve and increase alignment and collaboration with counterparts at all levels of government and within the private sector and regulatory bodies to create a holistic and integrated approach to data collection, sharing, analysis and use while maintaining public trust.

Statistics Canada's modernization agenda is built on five pillars which align with the agency's mission ("serving Canada with high-quality statistical information that matters"), vision ("delivering insight through data for a better Canada") and values (trustworthy, purposeful, caring and inclusive, and curious and always learning). These pillars provide the foundation for the Statistics Canada Data Strategy (SCDS). They respond to the ever-changing data landscape and to users' and stakeholders' requirements for more data, provided faster, and made available in multiple formats and from multiple access points. The modernization agenda aims to find the right balance between openness and transparency while protecting privacy and confidentiality.

Description: Statistics Canada Modernization

This image displays an overview of Statistics Canada modernization.  The five pillars of modernization are displayed as foundational blocks at the bottom of the image: user-centric service delivery, statistical capacity building and leadership, sharing and collaboration, leading-edge methods and data integration and flexible and modern workplace.

Drivers of modernization are displayed as icons at the top of the image: generation and collection, integration, dissemination, analysis and visualization, and consumption.

The goal of modernization is to delivery insight through data for a better Canada by providing better data, better insights and better services.

Statistics Canada's reputation as the country's premier, trusted data provider, and the agency's recognized expertise as a data steward, will further drive the modernization agenda toward an enterprise-wide (whole-of-government) approach to strategic data management with a focus on collaborative approaches to providing data and insights.

The SCDS has been organized along five strategic data capabilities and four sub-capabilities under data stewardship. The data stewardship capabilities provide the enabling infrastructure for achieving the other core capabilities. These five capabilities provide the foundation to ensure that the agency remains an independent and effective leader of the national statistical system, and a global leader in official statistics.

The five strategic data capabilities are:

• data governance
• data stewardship, including:
  • data discovery
  • data digitalization
  • data interoperability
  • data management
• data resources
• data trust framework
• data leadership

Each strategic data capability also includes short-term activities (what we are doing now), medium-term activities (what will we do next) and future activities (what will be the ideal mature state); as well as expected outcomes towards improved services to Canadians and increased evidence-informed decision making. Each strategic data capability will also require performance measurement criteria aligned to modernization. Performance criteria will be developed directly with each program area and will be integral to the next steps in the creation of a Data Stewardship Strategic Engagement Plan.

Activities at a glance

Strategic Data Capability	Short-term activities	Medium-term activities	Future state
Governance	Strategically realign to respond to ever-evolving data ecosystem and align with modernization agenda.	Formalize program data stewards. Increase intergovernmental leadership and synergies on data governance.	Develop well-established roles and responsibilities relating to enterprise-wide data governance and stewardship.
Discovery	Identify and increasingly gain access to administrative data sources for statistical purposes—master data are catalogued and available for use.	Establish and maintain ongoing, mutually beneficial relationships with data providers—ongoing data use monitoring offers BI insights to most-used and least-used data.	Collaborate with other government organizations (including the provinces, the territories and local area governments) to enhance the usefulness of administrative data for statistical purposes and reduce duplication.
Digitalization	Modernize IT infrastructure, ensure adoption of business-led and technology-enabled solutions, and ensure alignment.	Ensure alignment to IM vision, including moving to cloud-native storage and processing and role and policy based access control and data visualization.	Implement trusted digital identities and state of the art encryption. Implement data analytics as a service and remote data ingestion. Implement secure multi-party computing and privacy-preserving record sharing.
Interoperability	Research and select appropriate open data standards. Continue research and collaboration on international statistical standards and data reference models.	Continue to work on StatCan interoperability framework, ontologies, taxonomies, vocabularies and data visualization.	Conduct continuous research and improvement toward full interoperability.
Management	Create a "fit-for-use" StatCan data inventory. Design and perform initial seeding of statistical registers. Conduct a quality assurance framework (QAF) review.	Develop a fully realized corporate data inventory. Establish centralized and distributed data stores linked via virtualization layers. Develop realized statistical registers. Research enterprise QAF.	Ensure enterprise Government of Canada (GC) data inventory pulls directly from StatCan's meta store. Fully develop integrated statistical registers. Develop enterprise QAF (whole-of-government approach).
Resources	Establish programs to attract and retain employees with the right skills for current work, and ensure employees have appropriate tools to respond to agency priorities.	Conduct targeted hiring for employees with appropriate skill sets. Use refreshed suite of training options to develop and maintain required competencies and skills, including greater flexibility to access software and tools across the agency.	Conduct continuous campaigns to recruit and retain appropriate staff, and ensure evolutionary tool box for research and methods. Encourage and support culture of continuous learning to maintain relevant competencies and skills.
Trust framework	Ensure increased communication transparency of existing privacy and confidentiality practices and sharing of "root cause" user stories.	Research new and emerging methods for privacy and confidentiality protection. Implement role and policy based access control and automated auditing.	Fully automate data access control and audits. Implement appropriate levels of privacy control. Publish code and methods on Open Data Portal for transparency.
Leadership	StatCan, Privy Council Office and Treasury Board Secretariat (TBS) will clarify role of GC Chief Data Steward (CDS).	GC CDS, StatCan and TBS will co-lead work on the development of frameworks, principles, protocols and guidance for ethical and secure use of data.	Use continuous learning and GC and international information and best practices to exchange with other public and private sector organizations and academia.

Statistics Canada Data Strategy Framework

Description: Statistics Canada Data Strategy Framework

This placemat displays the Statistics Canada Data Strategy in the form of a summary framework.  The row headers provide a breakdown for each strategic data capability, governance, stewardship broken down by discovery, digitalization, interoperability and management, resources, trust framework and leadership.  The column headers provide information on evolution and progress sub-divided as short-term activities, medium-term activities and future activities.

The summary framework also includes a header with our pillars: user-centric service delivery, statistical capacity-building and leadership, leading-edge methods and data integration, sharing and collaboration, and modern and flexible workforce.

The framework contains a second header row with goals: people have knowledge and tools to access, use and share data, data are managed strategically through effective and horizontal governance and stewardship, data are accessible and secure, data are stored and accessed in an integrated and streamlined digital infrastructure and data literacy scientific methods and analytics are core competencies at StatCan.

The framework includes a footer row with our values: curious and always learning, caring and inclusive, trustworthy and purposeful.

Drivers, challenges and goals

Drivers

The greatest driver of the SCDS is the changing requirements for data and information of Canadians, Canadian businesses and Canadian institutions. Canadians want an authoritative source of information about what is important to them – relevant information and insights which respond to the increasingly complex economy and society. The legalization of cannabis, the opioid crisis, and the effect of foreign ownership on property values are some recent examples. Statistics Canada like many other National Statistical Offices (NSOs) around the world is responding to this need by reshaping its business model, building new networks and expertise, and devising new ways of unlocking the value of data for public good.

Challenges

The Data Strategy Roadmap for the Federal Public Service (DSFPS) included details on challenges faced by Government of Canada (GC) organizations:

• Absence of horizontal governance for strategic direction on data issues
• Lack of data literacy and cultural reticence to break silos
• Lack of adequate digital infrastructure and a complex rules framework
• Challenge of acquiring, governing, and managing large volumes of disparate data

Statistics Canada faces similar challenges. The agency modernization agenda is addressing these challenges through digital transformation in the form of IT modernization, the move to a broader use of administrative and alternative data sources, and fostering collaboration and partnerships with external partners including other NSOs.

Goals

The goals of the SCDS are to ensure that data are discoverable, accessible, interoperable, reusable, reproducible and open. These goals are realized through the following guiding principles:

• People have knowledge and tools to access, use and share data.
• Data are managed strategically through effective and horizontal governance and stewardship.
• Data are accessible and secure.
• Data are stored and accessed in an integrated and streamlined environment with modern digital infrastructure.
• Data literacy, a focus on modern scientific methods and analytical capabilities continue to be core competencies at Statistics Canada.

The SCDS is driven by the evolving and changing data and information requirements of Canadians, responds to challenges faced by GC organizations, including those detailed in the Data Strategy for the Federal Public Service, and will be realized through the activities listed under the key strategic data capabilities in this report.

Data governance

Statistics Canada's Data Strategy (SCDS) is aligned with, and in many cases supports or enables, the commitments of various federal government strategies such as the Data Strategy for the Federal Public Service (DSFPS). The SCDS ensures that Statistics Canada's data initiatives and priorities follow a common method and structure that is repeatable, thus ensuring efficient oversight, implementation, measurement and communication throughout the agency as well as the built-in potential to scale them across the enterprise if and when needed.

Statistics Canada's governance framework includes a policy suite (policies, directives, guidelines) and an organizational structure which includes accountabilities for its mandate and priorities. Statistics Canada's internal governance structure, with designated accountabilities for decision making around data, is documented in the roles and responsibilities sections of the Directive on the Management of Statistical Microdata Files and the Directive on the Management of Aggregate Statistics, and in the mandates of the senior management committees, including the Information Management Committee, the Microdata Access Management Committee, the Microdata Review and Release Committee, the Methods, Standards and Quality Committee, and the Communications and Dissemination Committee.
Statistics Canada Executive Management Board is divided into four committees, each with a clear mandate to facilitate responsible experimentation, innovation and action:

• the Strategic Management Committee
• the Modernization Management Committee
• the Corporate Planning Committee
• the Operations Committee

Statistics Canada has also designated several key executives with roles and accountabilities for data. These include Chief Information Officer, Chief Data Officer, Chief Technology Officer, Chief Financial Officer, Chief Procurement Officer, Chief Security Officer, Chief Privacy Officer and Chief Talent Officer.

Additionally, Statistics Canada seeks the expertise and counsel of many advisory groups and consultative committees, including the recently formed Advisory Council on the Modernization of Microdata Access and the Canadian Statistics Advisory Council whose mandate is to advise on the statistical system's overall quality, including the relevance, accuracy, accessibility and timeliness of its data.

Current agency governance practices ensure that agency programs and policies promote holistic data decisions and provide strategic direction aligned to modernization. Moving forward, Statistics Canada's governance structure must ensure that it is strategically aligned to better serve clients and citizens, and to reduce barriers and increase collaboration within the agency. Statistics Canada must work collectively to respond to initiatives and emerging needs, such as cloud adoption, microdata access, data science and engagement with Canadians. Creating synergies and collaborating horizontally across the agency in order to break down silos is vital. The organizational structure must be designed to turn vertical accountabilities into horizontal collaborations to ensure integration of cross-cutting activities. Statistics Canada's organizational structure must also ensure that roles and responsibilities include the following:

• conducting strategic planning to increase stakeholder engagement and outreach
• expanding the use of data visualizations, developing integrated portals, and creating horizontal storylines to promote responsive, comprehensive and collaborative analytical insights
• providing a full continuum of access from public dissemination to microdata integration across the agency
• optimizing emerging technological opportunities, including the continued digitalization of data to provide access to new data sources
• accessing and leveraging digital technologies such as the cloud, artificial intelligence (AI) and machine learning
• accessing powerful tools to effectively manage the volume, velocity and variety of data, including enhancing focus on data science, training and innovation
• focusing on cyber security and remaining vigilant to mitigate internal and external threats to the security of personal information
• enhancing responsiveness to emerging data requirements by recognizing the inherent horizontality of these issues and the agency's unique position as an integrator.

Short-term activities

• Realign and organize the agency in response to the evolving data ecosystem; this includes realigning senior management committees as needed.
• Formalize key executive responsibilities, including roles of CDO and CIO and supporting organizational units.
• Modernize Statistics Canada's Information Management (IM) Vision to ensure IM/IT alignment toward digital solutions.

Medium-term activities

• Formalize data stewards for each program.
• Increase intergovernmental leadership and synergies on data governance by co-chairing the Data Governance Working Group under the Enterprise Data Community of Practice.

Future activities

• Develop well-established roles and responsibilities relating to enterprise-wide data governance and stewardship.

Expected outcomes

• Increased consistency and confidence in decision making for requirements of the NSS.
• Better enabled planning and the minimizing or eliminating of redundant work and processes.
• Aligned accountabilities and performance measures for data and information.

Data stewardship

Data stewardship is the management and oversight of data assets to ensure they are of high quality, easily accessible and used appropriately in a consistent manner. Data stewardship is a business function that ensures alignment between business and information technology (IT) so that data assets are properly managed throughout their lifecycle. This includes ensuring that data are used efficiently and in a way that addresses privacy preservation, confidentiality and security requirements. Data stewardship is a key function of Statistics Canada and includes data discovery, digitalization, interoperability and management; these operational and tactical capabilities enable all other data capabilities that form part of the SCDS.

Data discovery

Statistics Canada is currently modernizing its statistical programs to respond to a rapidly changing and increasingly complex economy and society, a proliferation of data and providers, and increased user expectations for "real-time" and micro-detailed data

Statistics Canada's modernization includes an administrative-data-first agenda, an environment where data and insights already present in the data ecosystem are leveraged before surveys or other collection methods are used. This includes creating strategic partnerships with other organizations and researching and discovering data inputs that can be used by statistical programs, such as administrative data, open data, found data, commercial data, crowdsourced data and web-scraped data, while respecting privacy and maintaining public trust.

Statistics Canada is committed to the privacy, confidentiality and protection of data and information pertaining to Canadians, Canadian businesses and Canadian institutions. Current data acquisition research efforts have been directed at ensuring that Statistics Canada brings in the right data and that the data obtained is proportional and effective to the need for the data, having considered alternatives.

The agency must also ensure it continues to have the appropriate skills and tools to continue on the data discovery journey. Statistics Canada has responded to this challenge by developing tools, templates and communications to streamline the process, including an Administrative Data Inventory, an administrative data handbook, updated policy instruments and common templates for acquisition agreements.

The first phase of modernization at Statistics Canada included the implementation of four pathfinder projects which helped guide modernization, engage with stakeholders, consult with users, and implement leading-edge tools and methods to produce relevant, high-quality statistics. Over the last several years, efforts have been concentrated on discovering and acquiring administrative data files that directly feed the pathfinder projects: Measuring Cannabis, Transition to a Low-carbon Economy (Clean Technology), Measuring Growth in International Visitors to Canada, and the Canadian Housing Statistics Program. These pathfinder projects are providing lessons learned for the broader modernization plan at Statistics Canada. Specifically, lessons learned in acquiring these new data sources will be leveraged as the agency continues to pursue its administrative-data-first agenda.

Statistics Canada is also building capacity in the areas of negotiation and outreach, particularly as they relate to building strategic partnerships and win-win relationships with data and information providers. To ensure transparency, the agency is developing communications related to the acquisition, treatment and use of administrative data files by Statistics Canada.

Short-term activities

• Identify and increasingly gain access to administrative and alternative data sources for statistical purposes to produce more timely insights and to fill data gaps.
• Continue to implement, develop, and elaborate the administrative data first agenda and ensure employees have the necessary tools and skills for data discovery and access.
• Create clear, plain language communications around the acquisition, treatment and use of administrative data to ensure transparency, and work toward a more self-serve model for use by all Statistics Canada programs.
• Update the quality guidelines to reflect the increased use of administrative and alternative data in statistical programs.
• Update current policies and directives to respond to the administrative data first agenda.
• Create new policies and directives to respond to the requirement to bring alternative data sources into the agency, including the creation of a policy instrument and operational procedures for web scraping.
• Update policy instruments regarding the implementation of proportionality and necessity concepts during the data discovery stage.

Medium-term activities

• Establish and maintain ongoing, mutually beneficial relationships with data providers.
• Establish mechanisms to include community engagement in data acquisition and use, including using stories on how Statistics Canada's output leads to programs and policies that affect the lives of Canadians, and engaging in open dialogue on privacy concerns and data treatment at Statistics Canada (de-identification, anonymization, use of synthetic data).
• Collaborate with the Office of the Privacy Commissioner of Canada to finalize the development of the proportionality framework.

Future activities

• To foster interoperability from the local to the national level, collaborate with other government organizations (including the provinces, the territories and local area governments) to enhance the usefulness of administrative data for statistical purposes and to reduce duplication, including ensuring the optimization of the legal framework.
• Establish a culture of data reuse and sharing to ensure efficiency, especially with public sector data assets.
• Ensure Statistics Canada has well-established processes to support the acquisition of data that are shared, accessed or ingested from multiple access points.

Expected outcomes

• Data gaps are filled in order to provide new, relevant information to Canadians and support evidence-informed policies (ensure that funding is directed to where it's needed most).
• Increased transparency of the processes and uses of data to maintain social license and trust of Canadians.

Data digitalization

Data-driven organizations require a digital transformation plan. Under the SCDS, a key requirement of our digital transformation is to align operational decisions to the systematic (and automatic as much as possible) interpretation of data. This is rooted in Statistics Canada's ability to deploy advanced analytics at every point of interaction—human as well as machine—to continuously improve decision-making quality and accuracy. The SCDS also ensures that the agency adopts "business-led/technology-enabled" solutions.

The need for tools to collect, store, analyze, manage, share and visualize data is increasing in all departments. Enabling open standards, open source, interoperability and expertise sharing requires access to a common set of data tools commensurate with data needs. Departments also need a flexible framework to explore new tools and more advanced options that are both interoperable and secure. The Government of Canada's (GC) IT infrastructure must be able to support its data system's ambitious agenda. There is a growing need for higher computing capacity and for the modernization of older data infrastructure^{Footnote 1}

Statistics Canada continues to renew its aging and mission-critical IT infrastructure and systems. This work has been prioritized to enable innovative solutions that support modernization.

Statistics Canada's three-year IT strategy includes moving from a "custom-build" approach to a "borrow and buy" approach—only building when necessary to improve agility and responsiveness, and leveraging relationships and collaborating more closely with government departments, other national statistical agencies, the private sector and academia.

The following strategic actions will support the agency's modernization agenda and the GC's direction:

• implement a modern IT organizational model to better support the agency's strategic direction while ensuring sustained ongoing operations
• report regularly on key areas of IT system health performance to drive continuous improvement and reduce maintenance cost
• further develop IT skill sets and partnerships in the areas of IT security, open source, data science, cloud-related disciplines, agile development, user experience design, and technologies such as artificial intelligence and machine learning, smart machines, analytics, digital workplace platforms, containers, and micro services
• implement bimodal governance to better support innovation and remove unnecessary bureaucracy, while aligning and leveraging GC enterprise governance
• develop a digital strategy that aligns with Statistics Canada's modernization agenda and GC digital policy
• modernize IT operational services to support new technologies and approaches, such as cloud enablement and software-defined infrastructure
• adopt risk-based security approaches to support the cloud-first strategy, data access and a mobile workforce
• create a digital workplace so that all employees can work from anywhere, at any time, using modern secure tools
• facilitate greater access to Statistics Canada data holdings, including anonymized microdata, by giving researchers, policy makers and other stakeholders convenient virtual access to more data and statistical information that will be enriched by new dissemination tools, such as data visualization dashboards and portals, all while continuing to protect privacy and confidentiality
• offer powerful data analytics services with key components like a data management hub, data ingestion services, and data analysis using artificial intelligence, machine learning techniques and high-performance data processing.^{Footnote 2}

Short-term activities

• Deploy experimental projects to secure cloud in fall 2019.
• Begin Data Analytics as a Service (DAaaS) business requirements gathering.
• Develop business case and requirements for high powered computing (HPC) resources to facilitate scientific research (AI, ML, deep learning).

Medium-term activities

• Implement cloud native storage and processing and role and policy based access control and data visualization.

Future activities

• Implement trusted digital identities and state of the art encryption.
• Implement data analytics as a service and remote data ingestion.
• Implement secure multi-party computing and privacy-preserving record sharing.

Expected outcomes

• Digital projects and initiatives successfully align with the GC digital direction including enterprise-wide IT standards, security, and strategic direction ensuring consistency of approach and efficient and effective deployment.

Data interoperability

Interoperability is the ability to access and process data from multiple sources without losing meaning, and then integrate those data for mapping, visualization, and other forms of representation and analysis. Interoperability enables people to find, explore and understand the structure and content of datasets. In essence, it is the ability to "join up" data from different sources to help create more holistic and contextual information for simpler, and sometimes automated, analysis; better decision making; and accountability purposes.^{Footnote 3}

Adopting standards and adhering to them fosters interoperability, integration and rapid response. Statistics Canada's standards ecosystem includes data standards, classifications, foundational models, and vocabularies or ontologies.

Statistics Canada has experience and networks, nationally and internationally, to facilitate the definition of standards. National and international standards development organizations are modernizing methods for governing data. Canada is participating to address this still relatively new challenge and maximize our impact through a collaborative approach.

Statistics Canada actively develops and maintains statistical standards, promotes and monitors their implementation, and provides guidance on their interpretation. Statistical standards include populations, statistical units, variables and classifications (based on international standards to ensure interoperability).

Statistics Canada also develops, maintains and disseminates statistical metadata for statistical programs under the terms of the Policy on Informing Users of Data Quality and Methodology. The metadata include information on the variables, classifications, data sources, methodology, data quality, questionnaires, survey questions and response choices.

Future research will be directed toward the use and adoption of open data standards and information exchange protocols. This work will ensure interoperability across the government and with key stakeholders, including non-governmental organizations, think tanks, academia and private sector organizations.

Short-term activities

• Research and select appropriate open data standards.
• Continue research and collaboration on international statistical standards and the data reference model, including current commitments to HLG-MOS.
• Create vision, concept case and proof of concept for analytics data pipeline using common, industry-standard open data format.
• Continue participation in the Canadian Data Governance Standardization Collaborative.

Medium-term activities

• Continue to work on the interoperability framework: ontologies, taxonomies, vocabularies and data virtualization.
• Contribute to Pan-Canadian Real-world Health Data Network, metadata standardization.

Future activities

• Conduct continuous research and improvement toward full interoperability.
• Move Statistics Canada toward a standards accreditation body to create national standards.
• Implement Aria classification management (reference data) which forms the initial stages of a whole-of-government approach to classification management.

Expected outcomes

• Adherence to standards will optimize use of data and facilitate sharing and collaboration to ensure whole of government approach and ensure greater public value from data and greater usability and reusability of valuable data assets.

Data management

Data management is the practice of organizing and maintaining data processes to meet ongoing information lifecycle needs. Key components of data lifecycle management include a searchable data inventory, reference and master data management, and a quality assessment framework.

As a knowledge-based agency and a highly regarded National Statistical Office (NSO), Statistics Canada takes pride in the information it produces. Effective and efficient management of the data and information collected, compiled, analyzed and published is integral to meeting Statistics Canada's mandate. However, changes in technology have introduced new challenges in data and information management worldwide. Information is being produced at an ever-increasing rate because of the ease with which it can be created, transmitted and duplicated. New electronic formats continue to emerge, while others become obsolete. Organizations are finding themselves burdened with large amounts of unorganized legacy information to sort through while new information is being created at an unprecedented rate.

Increasing demands for flexibility in the work environment, the need to be agile when responding to evolving client and employee needs, and the radical technological shift to cloud are changing the information management landscape. There has never been a stronger need for an integrated, user-based approach—focusing on what information users need and in what form they need it—coupled with the provision of a trusted and authoritative data source. Statistics Canada must fully integrate information management into business processes and IT solution development to excel in and fulfill its mandate to provide insights to Canadians while meeting privacy preservation, confidentiality and security requirements.

Statistics Canada is developing a renewed, lasting and comprehensive enterprise-wide information management vision. This is a critical first step for Statistics Canada to respond to these new information challenges and to ensure the continued delivery of its products and services in an efficient and effective manner.

When fully deployed, the Statistics Canada Picasso platform will give all employees the ability to search and discover data assets and statistical metadata. Data assets include data from all sources, including surveys, administrative data, record linkage projects, big data and web scraping. The Picasso platform will be a key enabler of data stewardship of key strategic assets, as it delivers machine-readable metadata to drive processes, interpretability, coherence, confidentiality, replicability and reversibility. It will be a single toolkit for managing data and statistical metadata, and will lead to minimum storage of archived files. Picasso will replace several legacy systems and ensure a holistic approach to metadata standardization, access and use.

Reference and master data management are key components of enterprise data management. Statistics Canada has reviewed Scandinavian models for modernizing and creating statistical registers and has launched projects to research, develop, initialize, maintain, update and integrate an interoperable and flexible statistical register infrastructure that will serve as the foundation of statistical activities at Statistics Canada. This includes statistical registers or databases on:

• Buildings and building units
• Geospatial information
• Businesses
• Population

Data quality management and quality control are core business activities at Statistics Canada. A key commitment of the Data Strategy for the Federal Public Service is creating a comprehensive GC data quality framework. Statistics Canada and the Treasury Board Secretariat (TBS), under policy direction and with input from other selected organizations, will develop, implement and support data quality management strategies, policies and practices to support the framework. The proposed framework will establish common enterprise-wide standards to ensure interoperability and will be adaptable to the various realities and data ecosystems of GC organizations.

Statistics Canada currently has several strategies and tools in place to facilitate and ensure effective quality management in all its statistical programs and initiatives, including the Quality Secretariat, the Quality Assurance Framework (QAF) and Quality Guidelines. These tools can be adapted for an enterprise-wide approach to the quality and maintenance of GC data assets.

Short-term activities

• Support the administrative-data-first approach by identifying and developing processes and tools to improve user-centric services by establishing a strengthened self-serve model (Data Governance and Stewardship Support Services).
• Create clear and plain language communications about the use of the self-serve model (Data Governance and Stewardship Support Services).
• Perform Picasso enhancements that respond to needs of Data Stewardship, Data Analytics as a Service, and reporting requirements of approved specific performance measures.
• Design and perform initial seeding of Statistical Registers, and present concept case for reference data as a service (RDaaS) to the Government of Canada Enterprise Architecture Review Board (GC EARB).
• Work with TBS and other key business owners to establish government-wide master data management program.
• Conduct a quality assurance framework review related to administrative data and non-statistical uses of data and the creation of interdepartmental working group.
• Create improved data stewardship toolkit (training, documented processes and procedures, policy instruments, performance measures, etc.).
• Update the quality guidelines to document the proportionality and necessity concepts for data lifecycle management.
• Update policy instruments regarding the implementation of proportionality and necessity concepts for data lifecycle management.
• Co-chair the Data Quality Working Group under the Enterprise Data Community of Practice.

Medium-term activities

• Develop a fully realized corporate data inventory (Picasso).
• Further develop statistical registers and decommission other registers and frames.
• Research enterprise QAF, which includes assessing the applicability of Statistics Canada's QAF and Quality Guidelines for broader use, including non-statistical data uses; begin developing quality-related training strategies for the GC.

Future activities

• Ensure the enterprise GC data inventory pulls directly from Statistics Canada's meta-store (Picasso platform).
• Fully develop integrated statistical registers (data flow through registers for harvesting and de-identification).
• Develop enterprise QAF (whole-of-government approach).

Expected outcomes

• Data is used more efficiently and effectively throughout the enterprise to ensure everyone views and analyses the same, most recent information to ensure greater usability and public value from data.

Data resources

Statistics Canada currently has a robust recruitment program in place for analysts, IT professionals and methodologists. Recent recruitment campaigns have targeted data scientists, data engineers, business-intelligence experts and cyber-security professionals. Statistics Canada is also continuing to create strong development programs for all new recruits and for current employees seeking to update their skills or learn new ones. Statistics Canada strives to be a "learning organization" and is constantly building statistical capacity and increasing data literacy.

In a time of rapid change and increasing competition for talent, collaboration with partners outside Statistics Canada is essential in progressing the agency's modernization agenda and its ability to remain responsive to the needs of Canadians and decision makers. Partnerships with postsecondary institutions provide mutually beneficial opportunities with positive short-term and long-term outcomes, including:

• creating mutually beneficial learning opportunities
• creating practical student work experiences for the future workforce, and gaining access to new and developing talent
• establishing career opportunities for highly qualified, experienced graduates who are aware of and interested in the agency's work (this could be at several levels, such as undergraduate [recruit level] and postdoctoral [mid-level])

The Statistics Canada postsecondary institution partnership approach outlines a framework and assigns roles and responsibilities to the Strategic Management Committee, Modernization Management Committee, Talent Acquisition Committee, assistant chief statistician-level champions, director and assistant director partnership project leads, partnership project teams, and the Human Resources Branch.

Currently, the agency is actively recruiting individuals from the data science community, given the growing requirement for these types of skills within the agency and across all industries. Statistics Canada has developed a data scientist development strategy and partnership framework that outlines key success factors and expected outcomes.

In light of Statistics Canada's growing need for data scientists, an inter-field working group at the senior management level has been created to establish a corporate recruitment strategy for data scientists. To ensure the right talent and capacity are in place now and for the future, the working group will address human resources practices, including traditional and innovative hiring.

The agency recently created the Data Science Division, which will promote a hub and spoke model to increase awareness, provide expert advice to further data science projects, and expand the use of data science tools and techniques across the agency. Statistics Canada also recently launched a data science fellowship program. Data science fellows will be employed by Statistics Canada and will work within GC departments to leverage its wealth of data for insight and decision making.

Statistics Canada also has a growing number of grassroots communities of practice (CoPs), which self-manage and share a wealth of information and learning resources for data science and general data literacy skills. Current CoPs include the machine learning CoP, the data analytics CoP, and the geomatics CoP, as well as the R/Python Users Group, the Power BI Users Group and the Machine Learning Club.

The agency has also organized several innovation hackathon events, which are collaborative problem-solving events that bring together multidisciplinary teams (GC and academics) to build capacity to mobilize data; increase data literacy; and socialize leading-edge methods and tools, data stewardship best practices and user-centric approaches. These hackathons also provide an opportunity for the agency to attract new talent and encourage more collaborations.

Statistics Canada is also ensuring that, as resource requirements change, the agency provides employees with opportunities to grow and to learn new skills and capabilities to meet these changing requirements. As the agency moves toward more partnerships, more data sharing, and more ingestion of data assets from external sources, the requirements will grow for employees with partnering, negotiation, and engagement skills, and for technical writers, strategists and storytellers.

Short-term activities

• Understand desired competencies and skills for data literacy and data science, including artificial intelligence (AI) and machine learning.
• Establish science-based CoPs and working groups with access to appropriate tools; expand recruitment programs to attract and retain data scientists and data engineers.
• Work directly with the Canada School of Public Service (CSPS) on the Digital Academy roll-out and skill requirements.
• Launch the Data Science Fellowship.
• Establish a Data Science Division.
• Establish an Innovation Secretariat.
• Conduct continuous recruitment campaigns directed toward the right skills required for current priorities.
• Launch a results-driven research and development (R&D) board.
• Work directly with partners across the federal government to establish data literacy competencies for public servants, measure literacy and build training programs.

Medium-term activities

• Collaborate with the CSPS to refresh the suite of training offered to develop and maintain required competencies and skills.
• Increase targeted hiring (science-related skills), AI, machine learning, and the ethical use of automation and AI and other skills as required.
• Improve flexibility to access software and tools across the agency.
• Increase sharing and collaboration within and outside the GC, including access to and use of GitHub, GitLab, Trello and Slack, as well as GCconnex and GCcollab.
• Become the test bed for data science acceleration in the federal government; have GC data scientists work on Statistics Canada premises (with access to data, registers, quality assessment framework (QAF), analytical resources) to create analytical code and processes.

Future activities

• Encourage and support a culture of continuous learning to maintain relevant competencies and skills (Statistics Canada as a "learning organization").
• Run continuous campaigns to recruit and retain appropriate staff (flexi-CAD, free agent program, academic exchange).
• Establish an evolutionary toolbox for research and methods (ensure scientists have access to the right tools and resources to effectively and efficiently perform their work).
• Become a world leader in data science.

Expected outcomes

• The skills and competencies required to ensure data is used as a strategic asset, including data science, data analysis, communications, engagement and strategy are addressed and responded to ensure increased evidence-informed decision-making.

Data trust framework

The SCDS mission statement is "delivering insight through data for a better Canada." This is expressed in Statistics Canada's core values, which are to be trustworthy, purposeful, caring and inclusive, and curious and always learning.

Statistics Canada has spent more than a century collecting data for statistical purposes and is evolving as information needs grow and new data sources offer new possibilities for information. Currently, the agency is working with experts from businesses and academia from around the world, as well as with the Office of the Privacy Commissioner, to develop new statistical methods based on necessity and proportionality. Necessity determines what data Statistics Canada is mandated to produce to ensure society, the economy and the environment are measured. Proportionality ensures that the volume of data and data sources required by program areas is balanced with response burden, while maintaining protection of privacy. This work complements the work of the agency to ensure social acceptance by maintaining social licence, which is a key driver of the data trust framework.

Statistics Canada currently has, in addition to robust physical and IT security measures, a strong confidentiality awareness program and corporate tools for data access control and confidentiality classification; the application of the new proportionality framework further enhances the privacy preserving principles of the agency.

Future direction is aimed at increased communication and transparency of current practices, and research on new and emerging methods for role and policy based data access control, automated active auditing, and the development of protocols on appropriate use of synthetic, de-identified and anonymized datasets for training, testing and research as well as the development of improved privacy and confidentiality preserving methods.

Short-term activities

• Increase communication transparency of existing privacy and confidentiality practices.
• Work collaboratively with other National Statistical Offices, businesses, academia and the Office of the Privacy Commissioner to develop new statistical methods based on necessity and proportionality.

Medium-term activities

• Research new and emerging methods for privacy and confidentiality protection.
• Implement role and policy based access control and automated active auditing.

Future activities

• Fully automate data access control and audits.
• Implement appropriate levels of privacy control.
• Publish code and methods on the Open Data Portal for transparency.

Expected outcomes

• Protection of individuals' information and privacy ensure that Canadians remain confident of the work of Statistics Canada and the agency remains a trusted source of data and insights.

Data trust leadership

As a result of the strategic data capabilities outlined throughout this document, Statistics Canada is well positioned to take a leadership role in ensuring the strategic use of data across the federal government. Through partnerships and collaboration with all levels of government, the agency is working to optimize data access and use, provide guidance on data management, and support data-driven innovation in order to bring value to Canadians. The agency is in a position to inform on standards to ensure consistency across the federal government and to lead in areas such as the ethical and secure use of data, data quality, data catalogues and inventories, data integration, capacity building, and information generation through statistical science. Statistics Canada is also providing guidance and support to other departments on the development and operationalization of their departmental data strategies.

Statistics Canada co-led, in partnership with the Treasury Board Secretariat (TBS) and the Privy Council Office (PCO), the drafting of the Data Strategy Roadmap for the Federal Public Service (DSFPS). Statistics Canada is currently chairing or co-chairing several federal government working groups under the Enterprise Data Community of Practice, including the Data Governance Working Group, the Data Catalogue Working Group, the Data Reference Working Group and the Data Quality Working Group. The goal of the working groups is to build capacity throughout departments by sharing expertise and best practices. Statistics Canada's role is to share the agency's unique expertise in making relevant and high-quality data accessible to users while minimizing the risk of disclosing information on a specific individual, business or institution.

While Statistics Canada is leading, co-leading and supporting on a number of initiatives from the DSFPS, the agency is currently developing implementation plans for three key initiatives which have been identified as a priority for the agency: data literacy and training, data stewardship engagement strategy, and a data science community of practice.

The data literacy initiative is being launched in response to the need, as outlined in the DSFPS, for data literacy training to increase competency and skills among public servants. This need has also been identified by Statistics Canada's data users and stakeholders. The goal of the initiative is to provide user-centric training that targets a range of skill levels such that learners will have the necessary skills to use and understand Statistics Canada data to make evidence-based decisions.

The data stewardship engagement initiative will begin by focusing on the sharing of reference and master data. A pilot project, in collaboration with Employment and Social Development Canada (ESDC), will demonstrate how standardized reference data can facilitate interoperability and the use of data sharing platforms. The outcome of the pilot is the creation of a proof of concept prototype whereby reference data stewarded by Statistics Canada is accessed by other GC organizations thus releasing them from the burden of creating and stewarding the data themselves. The longer-term objectives of the data stewardship engagement strategy are to leverage partnerships and key governance tables to optimize the use of data as a strategic asset across the federal government by making high-quality data discoverable and accessible in a consistent manner, leveraging expertise across the enterprise, facilitating data sharing and integration, reducing duplication, increasing transparency, and ensuring that Canadians have greater trust in the quality, confidentiality, and value of public sector data and information.

The goal of the data science community of practice (CoP) initiative is to create an inter-governmental CoP for data science specialists. This CoP will establish the foundations of a government-wide data science ecosystem which will allow expertise to be shared across the government for establishment of best practices. The outcome of the CoP will be the creation of a forum for knowledge sharing and collaboration, GC capacity building in data science, and the creation of collaborative networks with academics and industry to further drive government outputs.

As a world-class National Statistical Office, Statistics Canada also occupies a leadership role on the international stage including within UNESCO (the United Nations Educational, Scientific and Cultural Organization), OECD (Organization for Economic Co-operation and Development), UNECE (United Nations Economic Commission for Europe), UNSD (United Nations Statistics Division) and the HLG-MOS (High-Level Group for the Modernization of Official Statistics), to name a few. National Statistical Offices around the world are showing leadership in governance, ethics, privacy protection, quality, data sharing, data interoperability, analytics, data stewardship, data literacy, and more. Statistics Canada represents the nation on international committees, associations and working groups on subjects such as data architecture, international standards and standards boards, data sharing and strategic partnership frameworks, data interoperability, and statistical methods and research. The agency achieved this by fostering collaboration through international partnerships. While Statistics Canada shares expertise in many of these forums, the agency also gains insight and information that is beneficial to the agency and National Statistical System. This is information that can be shared across the GC to further drive enterprise data management and stewardship.

Short-term activities

• With the Privy Council Office and TBS, clarify the potential role of Chief Data Steward for the Government of Canada.
• Develop work plans and implement on three key initiatives: data literacy training, data stewardship engagement strategy, and data science community of practice.
• Continue leading and expanding on innovative partnerships and experimentation (for example through hackathons and horizontal reviews).

Medium-term activities

• In partnership with TBS, co-lead work on the development of frameworks, principles, protocols and guidance for ethical and secure data use.

Future activities

• Develop modern and non-invasive stewardship rules, guidelines and principles that involve the changing data ecosystem through continuous learning and GC and international information and best practices exchange with other public and private sector organizations and academia.

Expected outcomes

• Optimized use of data across the federal government resulting in increased public good; increased public trust and confidence in the value of public sector data and information.

Next steps

Shortly after submission of the SCDS on September 30, 2019, a governance committee (new or existing) will be assigned responsibility for creation of an implementation plan for the SCDS. The governance committee will ensure that accountabilities for the SCDS, in the form of performance measures, are assigned to program areas. These program areas will work with the Strategic Data Management Branch and the Corporate Strategy and Management Field to ensure alignment to modernization and to the requirements of the 2019 Management Accountability Framework.

Statistics Canada is committed to sharing best practices, building capacity, collaborating with our partners, and meeting our users' needs as the data landscape evolves. The SCDS is a living, evergreen document that will be updated as required in response to this evolution.

Glossary of terms

Data ecosystem

A data ecosystem is defined as the collection of stakeholders, capacities, processes, policies and infrastructure used to capture and analyze data.

National statistical system

The national statistical system (NSS) is the ensemble of statistical organizations and units within a country that jointly collect, process and disseminate official statistics on behalf of national government. OECD Glossary of Statistical Terms, February 4, 2004.