Administrative data are information collected by government or private sector organizations as part of their ongoing operations. Examples include records of births and deaths, taxation records, records about the flow of goods and people across borders, and data collected by satellites.
Like most other statistical agencies, Statistics Canada uses administrative data in lieu of or to complement survey data, and to support statistical operations. Using administrative data responsibly means the agency is able to improve data quality, meet new and ongoing information needs, reduce data collection costs and save time for Canadians who respond to our surveys. Administrative data are especially helpful to obtain data pertaining to populations or topics that may be difficult or costly to obtain by survey. Statistics Canada does this in a transparent manner.
Authority to obtain administrative data
The Statistics Act grants us authority to obtain administrative data. Statistics Canada ensures:
- privacy by collecting only the information we need to produce statistics and research, and limiting access based on the need to know
- security by keeping data safe from unauthorised access and use
- confidentiality by not releasing information that could directly or indirectly identify individuals, households, or businesses
- transparency in all our processes.
Use of administrative data in statistical programs
Administrative data enable Statistics Canada to fill information needs about Canadian society, the economy and environment. The use of administrative data reduces respondent burden, and improves quality and timeliness of data.
Statistical uses and benefits of administrative data include:
- Determining what we need to measure and the best means to do so
- Reducing response burden through the use of existing data
- Processing data to meet the highest quality and accuracy standards
- Benchmarking to validate the quality and accuracy of complementary data.
For additional information, visit Statistics Canada's Survey Methods and Practices Manual.
Principles for the Statistical Use of Administrative Data
- To reduce the duplication of information collected from Canadians, and in accordance with the United Nations Fundamental Principles of Official Statistics and the Recommendation of the Council on Good Statistical Practice of the Organisation for Economic Co-operation and Development, Statistics Canada makes use of administrative data in the production of official statistics.
- The statistical uses of administrative data are kept functionally separate from the administrative uses of such data through strong legal, policy and organizational safeguards. In particular, the confidentiality of all identifiable administrative data obtained by Statistics Canada for statistical purposes is protected by the Statistics Act and the data are only accessed within Statistics Canada based on demonstrated needs.
- In collaboration with data providers, Statistics Canada uses its mandate under the Statistics Act to access administrative data for statistical purposes, and to influence the information collected by departments of government (including the provinces and territories), in order to enhance the usefulness of administrative data for statistical purposes and reduce duplication.
- Statistics Canada's objective in using administrative data for statistical purposes is to achieve a better overall balance among cost, respondent burden and data quality, where quality refers to the relevance, accuracy, timeliness, accessibility, interpretability, and coherence of statistical outputs.
- Administrative data are used for many different statistical purposes: replacing or complementing direct data collection to reduce costs and respondent burden; achieving efficiencies in statistical operations, such as the creation of survey frames, design of survey samples, imputation, estimation, and the measurement of the quality of other data; and developing and providing access to new data products (tabular or analytical), such as by integrating administrative data with other data.
- Statistics Canada carefully considers the impact on privacy in using identifiable administrative data for a purpose that was not envisaged at the time of original data collection, particularly when administrative data are to be linked to other data.
- Statistics Canada considers the possibility of using administrative data already available before undertaking new data collection activity.
- Statistics Canada establishes and maintains ongoing, mutually beneficial relationships with data providers, both to ensure the long-term supply, usefulness and quality of administrative data, and to identify and influence future changes to administrative data sources that may affect statistical uses.
- Statistics Canada assesses the fitness for use of administrative data sources and, in collaboration with data providers, identifies ways to improve the quality of the administrative data. In particular, the processes for collecting and producing the administrative data within the data provider's organization and the effects on the proposed statistical use must be well understood, and the quality of the administrative data must be documented.
- Statistics Canada is transparent in its use of administrative data. It communicates the benefits of reduced respondent burden from using data already collected as well as the measures taken to protect the confidentiality of the data.
Fit for use assessment of administrative data
Statistics Canada aims to develop, produce and disseminate accurate and timely information that represents its data users and stakeholders. Administrative data products are therefore assessed against established criteria, such as relevance, timeliness, accuracy, accessibility, interpretability and coherence, to determine statistical usability. Statistics Canada's Quality Assurance Framework provides an overview of the approach used to support statistical program areas as they strive to achieve an appropriate balance between the various dimensions of quality and related factors, including the evolving needs of users, costs and respondent burden.