An overview of the upcoming planned revisions to the Canadian System of Macroeconomic Accounts

Statistical revisions are carried out regularly in the Canadian System of Macroeconomic Accounts (CSMA) in order to incorporate the most current information from censuses, annual surveys, administrative statistics, public accounts, etc., and to implement improved estimation methods. These types of revisions are referred to as annual revisions and are generally restricted to the most recent three to four years of annual estimates.

Statistics Canada also conducts what are often referred to as “historical” or “comprehensive” revisions to the Canadian System of Macroeconomic Accounts. These revisions reflect new concepts, accounting treatment, classification systems or methods as specified by international accounting standards or due to on-going research at Statistics Canada. In the past, Statistics Canada conducted these types of revisions every 10 to 15 years, with major revisions taking place in 1986, 1997 and, most recently, in 2012.

Statistics Canada has determined that following the 2012 release, revisions relating to concepts, methods, classification systems and accounting standards will be conducted on a more frequent basis and will be rolled into the annual CSMA revision process. The rationale for this change is driven by two factors. The first is to ensure that the CSMA remain relevant and reflect the current state of the Canadian economy. The second is to ensure that the CSMA are internationally comparable. Most other G-20 nations are conducting comprehensive revisions with increased frequency, and the CSMA is aligning itself accordingly. While the revisions to CSMA concepts, methods and accounting treatment will occur more frequently, they will be smaller in scale to those undertaken in the past. It should also be noted that all revised data will be available to users, free of charge, via Statistics Canada’s website.

The result is that there are now potentially three components to any CSMA revision. The first component involves incorporating the latest CSMA benchmarks (such as estimates of gross fixed capital formation, household final consumption expenditure, gross operating surplus) that emanate from the national and provincial Input-Output Tables. The Canadian Input-Output Tables incorporate Statistics Canada’s highest quality data sources in a rigorous and detailed accounting framework. As such, they represent the most detailed and coherent accounting system for the structure of the Canadian economy and are considered the most accurate benchmarks on which to anchor estimates in the CSMA. The Input-Output Tables are generally available 30 months after the reference period. For example, with the first quarter release of 2013, the benchmark estimates originating from the 2010 Input-Output Tables are available for the rest of the CSMA. Generally, only one vintage of Input-Output Tables are produced, but from time to time they are revised, and those revised benchmarks are also incorporated into the CSMA.

The second component involves incorporating the latest annual and administrative data in the current years (for which IO benchmarks are not available). This includes data originating from annual business surveys (Annual Survey of Manufactures, Annual Financial Statistics Survey, etc.) and updated administrative data files used to produce estimates of wages and salaries and corporate operating surplus. These data are usually integrated for the year T-2. For example, with the first quarter release of 2013, the 2011 estimates would be impacted.

The third component includes adding revisions associated with the inclusion of new concepts, accounting standards, new detail, improved estimation techniques, updated classification systems, or updated methodologies. These are referred to as comprehensive revisions and often impact the entire CSMA time period.^{Note 1} As noted earlier, these types of revisions were historically undertaken every 10 to 15 years. Moving forward, the CSMA will be implementing comprehensive revisions on a more frequent basis and they will be rolled into the annual revision process.

To summarize, any CSMA revision process could involve three components:

1. Incorporation of the latest benchmarks from the Input-Output Tables;
2. Incorporation of the latest annual survey and administrative data
3. Incorporation of new concepts, methods, accounting standards and classification systems

In general, the first two types of revisions will be carried out each year and these revisions generally only impact the CSMA time series for years T-1, T-2 and T-3, while the third type of revision will occur less frequently (once every 2 to 3 years) but will generally impact the entire time series.

Upcoming revisions planned for 2013, 2014 and 2015

Over each of the next three years the following revisions will be introduced into the CSMA.

In 2013, the CSMA will introduce 2009 and 2010 benchmarks originating from the national and provincial Input-Output Tables, as well as the latest data from Statistics Canada’s various annual survey and administrative data programs. In addition, revised estimates of consumption of fixed capital, originating from the capital stock program, will be included for the period 1981 to 2012. Finally, a number of smaller statistical revisions to eliminate breaks in various detailed series will be introduced throughout the CSMA time period. For the most part, these are being introduced to align the time series with the 2009 benchmarks. These detailed revisions will not impact any aggregate macroeconomic indicators but will improve the quality of the detailed time series.

In 2014, in addition to incorporating the latest benchmarks from the Input-Output Tables (reference year 2011) and annual and administrative data, the CSMA will incorporate new public sector data resulting from the conversion of Canada’s government statistics to the International Monetary Fund’s (IMF) government financial statistics (GFS) compilation guide. GFS is the internationally accepted standard for reporting government financial statistics, which Statistics Canada has been working towards over the last number of years. The GFS data will be integrated into the CSMA in 2014, and may result in a revision to the government sector data (revenues, expenditures, assets and liabilities). These revisions will be the result of improved data collection techniques, changes in the way the data are reported (a move to accrual from a modified cash treatment) and changes in the classification of government transactions. The CSMA will introduce these data from 2008 onwards, but if warranted (if there are breaks in the time series), may also revise years prior to 2008.

Finally, in 2015, Statistics Canada will undertake a comprehensive revision to further align with international standards and achieve coherence with the U.S. National Income and Production Accounts (NIPA) and the European System of National Accounts (ESA), following their comprehensive revisions. Important changes to those standards are being made over the next few years, changing the treatment of financial services and pensions. Statistics Canada envisions making similar changes, ensuring that our accounts remain compatible. The CSMA will also incorporate revised benchmarks from the Input-Output Tables as well as the latest annual and administrative data.

It is important that users familiarize themselves with this new revision policy. Statistics Canada endeavours to communicate upcoming revisions well in advance so that users can understand the revisions and adapt systems, analysis and models to incorporate new estimates.

Statistics Canada will continue to use the publication Latest Developments in the System of National Accounts to advise users of upcoming changes.

---

Note

1. In the past, these were often referred to as ‘historical’ revisions. The CSMA now refers to these as ‘comprehensive’ revisions, to align with the terminology used by the U.S. Bureau of Economic Analysis in their National Income and Product Accounts (NIPA).