2023 submissions

Purpose: The purpose of this project is to better understand what causes firms to adopt new technology and the consequences it has on firms and workers. To do so, a microdata linkage will be established between firm-level surveys on technology adoption (Survey of Business Innovation and Strategy, Survey of Advanced Technology, and Survey of Digital Technology and Internet Use) and employer-employee database (Canadian Employer-Employee Dynamics Database) as well as other databases (Census of Population and data on union representation votes for Canadian firms).

This project can help better inform Canadians on technology adoption and its impacts on the economy and labour market. In addition, it will provide relevant evidence and information to the academic community and policy-makers, which helps support the development of policies and programs to promote equal technology adoption and diffusion among businesses so as to increase Canada’s competitiveness and the benefit of people living in Canada.

Output: The output of this project will include several analytical reports that address the following questions:

1. What are the main factors that drive a firm’s decision to adopt technology?
2. How different are the patterns of technology adoption by businesses owned by subpopulation groups such as women and immigrants? Do they experience additional hurdles for technology adoption?
3. What is the relationship between unionization and technology adoption? Do unions act as facilitator or inhibitor of technology adoption?
4. What are the impacts of technology adoption on firm performance?
5. What are the outcomes of technology adoption on workers such as job displacement, changes in wages and inequality etc.?

The analytical file, without identifiers, will be made available via Statistics Canada Secure Access Points (such as Research Data Centres), and access will be granted to Statistics Canada deemed employees following the standard approval process.bation standard.

Purpose: The overall objective of the ESDC Employee Wellness Survey is to assess conditions in the work environment at ESDC and inform strategies that meet the needs of employees and optimize their well-being.

The purpose of the linkage is to add two variables to the ESDC EWS share file, which would be used to subset the data by Level of supervision and by official language. This would allow for analysis of principal survey results that would provide for a more in-depth analysis of these subgroups of respondents’ potentially different experiences to be understood and addressed in the form of improved people management practices.

Output: The planned outputs are a ESDC EWS Share file, and non-confidential aggregate statistics in the form of Excel tables and a Power BI dashboard, for Employment and Social Development Canada (ESDC). Statistics Canada will enter into a data sharing agreement with ESDC who in signing the agreement, agrees to keep the information shared confidential, and only use it for statistical and research purposes. Respondents to the ESDC EWS were informed of the sharing with ESDC at the time of collection, and only those respondents that agreed to share their information will be included in the ESDC EWS Share file. No direct identifiers, including personal identifiers, will be included on the ESDC EWS Share file. The ESDC EWS Master file placed in the Research Data Centres (RDCs) will not include the two linked variables. Only non-confidential aggregate statistics will be released outside of Statistics Canada.

Purpose: Building on the purpose of the 008-2018 linkage project, which was to build the capacity for identifying the primary risk factors and the sub-populations at greatest risk of an overdose. To create a better understanding of the characteristics of those individuals at the heart of the opioid crisis-particularly for those individuals using and dying in their residence. To aid in the effort to understand the roots of the illicit drug epidemic and the individuals most at risk of overdose. In addition to the policy perspective, if successful, synthetically generated opioid data can be used by researchers, health-care developers and clinical scientists to develop innovative health-care solutions and use it for teaching and training purposes.

This new project will utilize the same referenced cohort (008-2018 linkage project) to produce a generative Machine-Learning model for generation of privacy-enhancing synthetic datasets. Several Machine Learning models will be assessed to identify one which optimally balances privacy risks disclosures with data utility. Development and assessments of models and synthetic datasets will be a collaborative work between Statistics Canada and UQAM University researchers.

In addition, should the proof-of-concept be successful in balancing privacy and confidentiality risks against the data utility, it will allow useable privacy-enhancing granular-level synthetic data and study outcomes to a wider group of researchers and policymakers could encourage innovation through active collaboration and facilitate a broader and faster advancement of solutions to the opioid crisis. Synthetic patient data that preserves the relationship among study variables but contains no records that represents or identifies an actual individual in the cohort would be a viable solution to this problem.

Output: A comprehensive technical report summarizing the methodology, assessment of the generative algorithms, key findings, lessons learned and recommendations for next steps (if any). High-level findings may be reported in the form of presentations to various Public Safety Canada partners. Deemed employees of Statistics Canada will only have access to the data with an anonymized linkage ID, but NOT the direct identifiers, and use only authorised devices from Statistics Canada secure access points during this project.

A well-documented code repository for the project under Statistics Canada’s existing and future policies. As part of Open Science initiative, free access to the open-source tool and libraires will be rendered to public. Code will not contain sensitive information and will undergo appropriate assessments before release.

A pre-trained generative model that can produce a high-quality data in a differentially private setting. Such an approach in production could guide the development of targeted approaches for prevention, treatment, and identification of possible intervention points for the high-risk population in opioid-toxicity studies. This model will be capable of generating novel synthetic data instances not found in the original dataset which maintains the privacy of the members of the original dataset, while maintaining key properties that respect the data distribution.

No confidential Statistics Canada micro-data will be made publicly available during or after the completion of the research collaboration under this agreement. This term also extends to Machine Learning (pre-trained) models and prototypes that may in turn divulge confidential information.

Purpose: Relatively little analysis has been undertaken to measure farm-level productivity in Canada. This work will examine the degree that productivity growth is driven by improvements within continuing farms compared to how much results from the reallocation of resources like land between farms. This serves to inform policy aimed at improving the productivity of the agriculture sector.

Output: Only non-confidential aggregate statistical outputs and analyses that conform to the confidentiality provisions of the Statistics Act will be released outside of Statistics Canada. The information will be presented in the form of tables of regression results and summary statistics related to the project’s goal.

Analytical datasets will be placed in the Research Data Centres (RDCs) and access will be granted following the standard RDC approval process. The source datasets will be anonymized and will respect variable restrictions in effect for the source datasets. Access to the analytical file is restricted to researchers who have become deemed employees of Statistics Canada.

Purpose: The purpose of this linkage is to respond to the data needs of the Government of Canada’s Multilateral Framework for Early Learning and Child Care. This framework identifies key priorities for child care, including child care that is inclusive and flexible.

This microdata linkage will augment the 2022 Survey of Before and After School Care in Canada with information on income and employment characteristics, family structure and immigrant status in order to explore more fully characteristics associated with the use of child care in Canada.

Output: A linked microdata file will be available within Statistics Canada and will be placed in the Research Data Centres (RDCs) where access will be granted following the standard RDC approval process. Aggregate findings will be reported in research papers, internal and external reporting documents, presentations at workshops and conferences, as well as external publications (e.g., academic manuscripts).