Government Finance Statistics (GFS)

Detailed information for second quarter 2017





Record number:


The objective of this program is the publication of financial statistics concerning the federal government, the provincial, territorial, and local governments and the Canada and Quebec pension plans, based on the Government Finance Statistics (GFS) accounting framework developed by the International Monetary Fund (IMF).

Data release - September 21, 2017


The Government Finance Statistics program is designed to measure and analyze the economic dimensions of the public sector of Canada. These economic dimensions are: revenues, expenditures and the resulting surplus or deficit, assets and liabilities and net worth or net debt position.

Since government financial statements and reports are based on the organizational structures and on the accounting and reporting practices of individual governments, there is a lack of consistency across jurisdictions and over time.

Given this lack of uniformity in reporting practices, Statistics Canada (STC) over the last 65 years, in cooperation with representatives of all levels of government and with the academic and business communities, developed the Financial Management System (FMS).

The FMS was founded on a modified-cash based system of accounting. In recent years, Canadian governments have moved from a modified-cash based accounting system to an accrual based accounting system. As such the statistical system underlying government finance statistics must also change. Statistics Canada has decided to move towards reporting government finance statistics on a Government Finance Statistics 2001 basis. The GFS 2001 is an international accepted accrual accounting framework for government finance statistics. The GFS 2001 is also fully integrated with the United Nation's System of National Accounts framework.

The present statistical program replaces the descriptions in Consolidated Government Financial Assets and Liabilities (record no. 1709) and Consolidated Government Revenue and Expenditures (record no. 1735) became inactive after the publication of 2008 data.

While it will take a number of years to be able to derive detailed GFS-based statistics directly from government financial information (Statistics Canada will begin publishing public sector statistics based on the GFS 2001 manual in calendar year 2014), Statistics Canada has decided to release quarterly GFS data using Canadian System of National Accounts (CSNA) government sector data and a bridging model that maps these data to the GFS framework. The CSNA (consistent with the United Nations' System) already compiles some government data on an accrual basis and therefore offers the foundation to produce preliminary estimates of government data on a GFS basis.

These statistics are used in two broad ways. They provide a measure of the financial position of public sector components and sub-components. These statistical measures are used by a wide variety of economists and industry analysts in both the private and government sectors. Secondly, these data are used as the benchmark for the quarterly estimates of the Government Sector in the Canadian System of National Accounts (CSNA).

The core of the GFS analytic framework is a set of four financial statements. Each of these statements is described in the "Documentation" section at the bottom of this page.

Statistical activity

The Canadian System of Macroeconomic Accounts (CSMA) provides a conceptually integrated framework of statistics for studying the state and behaviour of the Canadian economy. The accounts are centered on the measurement of activities associated with production of goods and services, the sales of goods and services in final markets, the supporting financial transactions, and the resulting wealth positions.

Reference period: Quarterly data

Collection period: 2 months after the reference period


  • Balance sheets
  • Economic accounts
  • Financial and wealth accounts
  • Government
  • Government financial statistics
  • Revenue and expenditures

Data sources and methodology

Target population

The target population consists of all institutional units controlled and mainly financed by governments (federal, provincial, territorial, and local) in Canada except those classified as government business enterprises. The population covers all of the government components of the Public Sector Universe (PSU). The Public Sector Universe includes governments (federal, provincial, territorial, and local) and the Canada and Quebec pension plans. The government component includes all ministries, departments, agencies, non autonomous funds and organizations, universities and colleges, health and social service institutions and school boards.

Institutional units are comparable to enterprises in the Statistics Canada hierarchical structure of business units. Institutional units are economic entities that are capable, in their own right, of owning assets, incurring liabilities, and engaging in economic activities and transactions with other entities. Control may take the form of full ownership of the institutional unit or a majority holding of the voting shares. The availability of a complete set of annual financial statements is a prerequisite in order for an entity to be classified as an institutional unit within the government component of the Public Sector Universe.

Instrument design

Data for the federal government, as well as for the Canada pension plans (CPP) and the Quebec pension plans (QPP), are obtained entirely from administrative data sources. Data for provincial/territorial governments are obtained from administrative data sources and by surveys conducted by the Canadian Institute for Health Information (CIHI), and by Statistics Canada's surveys of residential care facilities and on school boards, colleges, and universities.


This survey is a census.

Data are obtained from a census of institutional units for all government levels in Canada as defined by the Public Sector Universe. The PSU contains all institutional units controlled and mainly financed by government. The PSU is maintained up-to-date through the public accounts and web sites of federal and provincial/territorial governments in Canada. Local government data are maintained from the administrative records of their respective provincial and territorial Departments of Municipal Affairs, from information contained in official provincial and territorial Gazettes, from municipal directories and from responses to on-going sub-annual municipal surveys.

Data sources

Responding to this survey is mandatory.

Data are extracted from administrative files and derived from other Statistics Canada surveys and/or other sources.

The data are compiled for the entire government sector population. This is made possible by utilizing publicly available audited financial statements, public accounts, and other administrative information available from federal, provincial, territorial, and local governments and their agencies. Data on federal and provincial/territorial governments as well as the Canada and Quebec Pension plans (CPP and QPP) are obtained entirely from administrative data sources. This administrative information is supplemented by financial details provided directly by the federal, provincial, and territorial governments, and in the case of local governments, by financial details obtained from questionnaires.

The concepts and definitions for most federal, provincial, territorial, and local governments are based on the guidelines of the Public Sector Accounting Board (PSAB) of the Canadian Institute of Chartered Accountants (CICA). Accounting practices are in accordance with the Generally Accepted Accounting Principles (GAAP) of the CICA.

Error detection

Most of the data come from audited financial statements of governments; therefore, minimal error detection procedures are required. For survey data, which represent roughly 1% of the total value, several automated checks are performed on the data to verify internal consistency and identify extreme values.


For non-response units, imputation is performed using historical information where historical information is available; otherwise, donor imputation is used. The donor imputation procedure involves using available auxiliary information to substitute the data from an entity with similar characteristics. Total salaries (aggregated to the relevant institutional unit from the T-4 form 'Statement of Remuneration Paid' from Canada Revenue Agency (CRA)) are used to benchmark the total expenses and imputation is done accordingly.

The coverage of the public sector population is virtually complete. Imputation for non-response varies by public sector sub-component, but for all components, the imputation rate is less than 2%. Similarly, the overall impact of imputation on major financial variables is also less than 2%.


Estimates are derived from the compilation of data obtained from the data sources for each institutional unit in the population of interest. The following processes are used to optimize accuracy:

A. Getting the detail:
Published public accounts and local government financial statements do not always contain the detail needed to precisely convert public accounts entries required for the CSNA classification. Generally speaking, the greater the detail in the source data, the greater the precision in applying classification codes. The practice is to first obtain the public accounts and then to approach individual governments and solicit the additional detail required to accurately apply the classification.

B. Quality control on processing:
Once public accounts publications and other financial statements are obtained and combined with supplementary information, there are many transactions required to transform these raw data into CSNA estimates. Strict quality control is maintained on all of these transactions such as historical continuity, data validation, and data confrontation.

Quality evaluation

The analysis of data that occurs before publication includes a detailed review of the individual responses (especially for the largest institutional units), a review of general economic conditions as well as historic trends and comparisons with original public accounts data before the conversion to the CSNA. Any anomaly is verified and resolved before data are published. An example of this cross-check occurs in the annual benchmarking of the government sector data with the Canadian System of National Accounts via the Input-Output Tables and the Gross Domestic Product series. The relevance of government finance statistics for the other parts of CSNA derives from the fact that governments are very large players in the economy whose financial transactions have to be included in the national accounts like any other large sector.

Differences exist between CSNA and pure GFS statistics. Government CSNA data, as noted above, originate from public accounts information. This information is adapted to the CSNA framework and is incorporated into products such as the Income and Expenditure Accounts, Financial Flows, National Balance Sheet and Input-Output tables. Finally, a mapping between the CSNA and GFS is performed to arrive at those quarterly estimates. In the mapping to GFS certain conceptual adjustments are made to the CSNA data (it should be noted that once the GFS is implemented, the data will be consistent with the Canadian System of National Accounts and the conceptual differences between the two sets of data will disappear). These conceptual adjustments are listed in the "Additional documentation" link at the end of this section.

The CSNA-mapped quarterly GFS data is Statistics Canada first step in providing government data consistent with the GFS framework. With the upcoming historical revision of the CSNA, to be published in 2012, the framework will be more fully implemented. In addition, the government data contained in the CSNA will be consistent with the GFS framework.

The detailed categories as well as the CSNA-to-GFS bridge table will be provided in Appendix A and a glossary of terms in Appendix B, contained in the article "Moving from the Financial Management System (FMS) to Government Finance Statistics (GFS)" in the publication "Latest Developments in the Canadian Economic Accounts" (Catalogue no.: 13-605-XIE). For ease of use, the IMF categories description and code structure is presented alongside the CSNA category description, as well as its Statistics Canada's socio-economic database (CANSIM) series identifier.

Disclosure control

Statistics Canada is prohibited by law from releasing any information it collects which could identify any person, business, or organization, unless consent has been given by the respondent or as permitted by the Statistics Act. Various confidentiality rules are applied to all data that are released or published to prevent the publication or disclosure of any information deemed confidential. If necessary, data are suppressed to prevent direct or residual disclosure of identifiable data.

Revisions and seasonal adjustment

The input data to the CSNA are not final until several years after the reference year, and the more recent the input data are, the more they are subject to revision.

In the case of government based statistics, data for the most recent year (for example year t) are based primarily on monthly information from the Government of Canada's banking and accounting system for the federal government; from budget forecasts and quarterly input from provinces and territories; from data estimated from previous years for education and health institutions; and from survey estimates from local governments. Data for the preceding year (year t-1) are primarily based on public accounts and audited financial statements for the federal government; on budget forecasts, quarterly input, and some public accounts data for the provincial/territorial governments; a mixture of previous years' estimates and survey data for the education and health institutions; and on survey data for local government. Data for the two preceding years (year t-2 and year t-3), for federal, provinces/territories and local governments, data are based on public accounts and/or audited financial statements and education and health institutions data are derived from survey results.

The 'fourth' preceding year (year t-4), corresponds to the benchmarking and revision cycle of the CSNA where all data from the CSNA are benchmarked to the Input/Output matrices and resulting Gross Domestic Product.

While the more recent data are necessarily more subject to revision than data for earlier years, the use of preliminary information (budget forecasts or survey for local governments) results in major advances in timeliness. Data are now released within three months of the end of the reference period. In light of the contribution of timeliness to the relevance of the data, this trade-off is in the interests of the data users.

Given that first estimates are based on budget forecasts (or survey estimates for local government) and that final data are based on public accounts/audited financial statements, the size of the revisions varies accordingly to the change from forecasts to actual data.

Data accuracy

The data produced are derived from a multitude of entities in the government component of the Public Sector. Statistics Canada has no control over the accuracy of the input data at the time they are received, although it does have the advantage of eventually having access to audited financial documents. We ensure that no errors are introduced through automated checks that verify internal consistency and identify extreme values, and we apply procedures that maximise the error-detection possibilities inherent in the data.

The inherent quality of the input data varies systematically through time, with the most recent data (current year) being the least reliable (and the least detailed) since they are largely based on government budget forecasts. For earlier reference years, with each additional year the input data becomes subject to smaller revisions. The public accounts and local government financial statements are eventually subject to audit and these audited accounts and statements form the benchmarks of historical data.


Date modified: