June 17, 2015
In 2010, Statistics Canada launched the Integrated Business Statistics Program (IBSP) to provide a more efficient model for producing economic statistics. The main objective was to enhance the economic statistics program so that it remains as robust and flexible as possible while reducing the burden on business respondents.
The IBSP encompasses around 60 surveys covering four major sectors: manufacturing, wholesale and retail trade, services (including culture) and capital expenditures. By 2019/2020, the IBSP will include roughly 150 economic surveys covering all sectors of economic statistics.
The program changes ensure that Statistics Canada will continue to produce a consistent and coherent set of economic statistics. As well, data users and researchers can more easily combine economic data with information from other sources to undertake their analyses.
The IBSP uses a standardized approach for economic surveys conducted at Statistics Canada. This framework involves:
- using a common Business Register as the unique frame
- maximizing the use of administrative information to reduce business response burden
- using electronic questionnaires as the principal mode of collection
- harmonizing concepts and questionnaire content
- adopting common sampling, collection and processing methodologies.
What are some of the more significant changes?
- A new approach to sampling ensures businesses will only be asked those questions that are pertinent to their operations. This creates a win-win situation for Statistics Canada and respondents. Statistics Canada reduces the collection effort and has a greater likelihood of collecting the information it requires to produce official statistics relevant to Canadians. It also reduces the time needed by respondents to complete their business surveys.
- Increased use of administrative data reduces business response burden. Administrative data files (such as corporate income tax files) have been used extensively as a direct substitute for a sub-set of sampled units and for imputation of non-response. In the transition to the IBSP model, imputation methods were adapted to take full advantage of the availability of administrative data. This resulted in additional response burden reductions across survey programs. The majority of sampled businesses are no longer required to provide data for revenue and expense information that is available from tax data. The IBSP questionnaires are designed to collect information that is not available from administrative data files, such as commodities produced and business practices.
- A new coherent approach to developing provincial/territorial estimates uses existing information on Statistics Canada's business register to determine provincial/territorial shares of revenues, expenses and value added. This ensures a coherent and standardized approach that is consistent across all IBSP surveys. Previously, these data were collected directly from each respondent, contributing to response burden.
- Electronic questionnaires are now the primary mode to collect data from business respondents. Businesses complete surveys using a secure online application. The result is a more efficient and higher quality collection process. In addition, the quality of survey statistics may improve because electronic questionnaires have built-in checks designed to limit reporting errors that can occur with paper-based questionnaires.
- Increased coverage of the business population results in a more comprehensive set of business statistics. Beginning in reference year 2013, the population covered by the suite of annual economic survey programs increased to include all firms regardless of their size. In previous years, relatively small businesses (based on their sales) were not included in Statistics Canada's central business frame. However, with new self-coding technology, it became possible to classify all businesses operating in the Canadian economy onto the central business frame, regardless of the sales of the firm. As a result, with improved coverage of the population, the IBSP-based estimates better reflect the population of businesses operating in Canada.
- Questionnaires have been updated to reflect the latest business terminology and accounting practices of Canadian businesses. In addition the questionnaires apply the latest standard classifications used by Statistics Canada, such as the North American Industry Classification System and the North American Product Classification System.
Does this impact the comparability of data through time?
The extent of the changes in the business statistics program introduced by the IBSP means that some series may no longer be consistent with estimates from previous periods. For example, the increase in the business population alone means that the estimates will tend to be higher than those previously published.
For some series, the 2013 changes will be small and comparisons with estimates for 2012 will be consistent. In other cases, the impacts can be significant, leading to breaks in the 2013 estimates data when compared to 2012.
Recognizing the importance of data continuity, Statistics Canada analyzed the 2013 estimates in comparison with 2012 data to determine whether a break in series occurred. Assessment techniques included:
- evaluating survey estimates at all levels of detail (national, sub-national, NAICS)
- comparing estimates obtained from sub-annual surveys (where applicable)
- comparing tax information
- analysing the results for common respondents in 2012 and 2013
- comparing historical movements by respondent and by the industry in general.
A break in series from 2012 to 2013 was identified when the direction and magnitude of the change for a given variable/province or territory/industry fell outside a survey's specific tolerance limit. The tolerance limit is defined in part as the bound between the largest increase and largest decrease using a forecast model. The tolerance limit could be further influenced by comparisons with auxiliary confrontation data, such as those noted above.
The determination of a series break focuses on the main aggregate variables only. These variables include Total Revenue, Total Operating Revenue, Total Expenses, Total Operating Expenses, Salaries and Wages, and Depreciation. Given the nature and scope of the changes to the survey program, series breaks for more detailed variables are inevitable and will not be analyzed by the agency.
In all cases, users are aware that breaks can exist and that any comparisons with the 2012 data should be made at their own discretion.
Once the estimates for the reference year 2014 are available, revisions will be made to the 2013 data as is normally the case. At that time, the 2012 estimates may also be revised due to the additional information available.
Who will use the new IBSP estimates?
- Businesses use the estimates to better understand their performance within their given industry relative to the industry average.
- Industry analysts use the IBSP estimates to analyze the performance of given industries in the Canadian economy both nationally and regionally.
The IBSP data are a main input in the Canadian System of Macroeconomic Accounts. They are first adjusted to macroeconomic accounting concepts and definitions and are then integrated into the macroeconomic accounting frameworks. This integration involves adjusting the data to adhere to the macroeconomic accounting identities as well as ensuring consistency through time. These data are the building blocks for Statistics Canada's benchmark measure of gross domestic product and a key input into the estimates used to determine equalization payments and the allocation of harmonized sales tax revenue.
Periodically, Statistics Canada undertakes large scale changes as part of its survey renewal process. The new IBSP data will be integrated into the Macroeconomic Accounts. Although the new data may lead to some changes/revisions to the national accounts, the System of National Accounts framework ensures that the national account estimates are robust and coherent.
To implement this important initiative, Statistics Canada is taking all the necessary steps to complete the final data and system verifications. The annual economic statistics are usually available approximately 15 months after the reference period, but the major transformation produced by the IBSP made it impossible to maintain this release schedule. Statistics Canada is committed to releasing the data as soon as possible, once all quality assurance and confidentiality checks have been completed.
It is expected that starting with reference year 2014, the release schedule will revert to respecting the 15 month timeliness objective.
For a more detailed explanation of the changes, please consult the Integrated Business Statistics Program Overview on the Statistics Canada website.
More information on the technical aspects on sampling and estimation is available upon request.
For more information, contact Media Relations (613-951-4636); work.wcms.statcan.ca/eng/node/17725/edit .