Canada Emergency Wage Subsidy Regional and Community-level Database

1. Overview

1.1 What is the Canada Emergency Wage Subsidy?

The Canada Emergency Wage Subsidy (CEWS) was a wage subsidy program that was implemented by the Government of Canada in response to the COVID-19 global pandemic. It was initially available for a period of twelve weeks, from March 15, 2020 to June 6, 2020, and provided a subsidy of up to 75% of eligible remuneration, paid by an eligible entity (eligible employer) that qualifies, to an eligible employee – up to a maximum of $847 per week.^Note  The CEWS program was subsequently extended. However, for periods starting from October 24, 2020, the subsidy amount was amended to be up to 55% of eligible remuneration, to a maximum of $573 to $595 per week per employee, depending on the period applied for.^Note  The CEWS program was administered by the Canada Revenue Agency (CRA) and was ended after 21 claim periods, on October 23, 2021.^Note 

1.2 What is the Canada Emergency Wage Subsidy Regional and Community-level Database?

The Canada Emergency Wage Subsidy Regional and Community-level Database Version 3.0 (the database) is a custom dataset constructed with CEWS microdata, CEWS aggregate data on the CRA website, and other administrative data sources available at Statistics Canada. In terms of geographic levels, the database provides values for Canada, province/territory, rural and urban breakdowns by province/territory, census metropolitan area (CMA), census agglomeration (CA), and census subdivision (CSD). These values are provided at the total industry level and by 2- and 3-digit North American Industry Classification System (NAICS^Note  2017 version 3.0) industries. The main variables included in the database are total CEWS supported employment and total CEWS claim dollar amounts. To generate this regional and community-level database, CEWS microdata were first aggregated to the business enterprise level, and subsequently allocated to the business locations of the enterprise. These granular data can shed light on the usage of the CEWS program and provide economic insight into the business performance of various industries in local communities throughout the country.

2. Data sources

2.1 Canada Revenue Agency CEWS microdata file

The CEWS program is organized by CEWS claim periods set up by the CRA. A CEWS claim period is the four-week period for which an eligible employer can claim the wage subsidy for remuneration paid to eligible employees.^Note 

The CEWS Version 3.0 database records cover the first fifteen CEWS periods:^Note 

• The first eight CEWS periods are based on a microdata file provided by the CRA containing all approved CEWS claims as of January 11, 2021 and are republished from the first version of this database (CEWS Version 1.0) released on March 18, 2021.
• CEWS periods 9 to 11 are based on a microdata file provided by the CRA containing all approved CEWS claims as of June 7, 2021.
• CEWS periods 12 to 15 are based on a microdata file provided by the CRA containing all approved CEWS claims as of November 8, 2021.

The CRA microdata files contain records for all CEWS claims approved at the time when the microdata file was produced. CEWS applications undergoing reassessment by the CRA after initial approval are excluded in subsequent files pending final approval status. It is possible for a business to make a CEWS claim up to 180 days after a four-week claim period has ended.^Note  As a result, the total number of CEWS-approved claims for a given CEWS period contained in CRA CEWS microdata files received at Statistics Canada can fluctuate depending on the date the microdata file is produced.

Version 3.0 of the database also uses CEWS data at higher levels of geographic (province/territory) and industry aggregation (2-digit NAICS codes) obtained from the CRA website.^Note  These data, contained in Version 3.0 of the database, were updated on the CRA website on February 8, 2022.

2.2 Statistics Canada’s Business Register

The Business Register (BR) contains a comprehensive and up-to-date list of businesses in Canada. It contains business information such as address, business number (BN), a unique operating entity number (OEN), among many other variables.^Note  This information was used for allocating CEWS variables to business locations, and assigning census subdivision geographies and NAICS industry codes.

3. Reference period of the database

This release of the database includes fifteen four-week claim periods, beginning March 15, 2020 and ending May 8, 2021. These fifteen periods are individually identified within the database.

The records for the first eleven CEWS periods spanning March 15 to January 16, 2021 released in Version 1.0 and Version 2.0, while included in Version 3.0, are not updated.

3.1 Determination of CEWS reference periods to include in the database

The November 8, 2021 CRA microdata file that was used to produce Version 3.0 of this database contains values up to October 23, 2021 (CEWS period 21). As per Section 2.1, microdata files are updated bi-weekly by the CRA and include data both for the new period as well as revisions of claims (e.g., due to late filings) for past CEWS periods. In general, the most recent CEWS periods exhibit larger divergences between files produced at different times and are expected to have a greater degree of revisions than earlier periods. As a result, this database publishes data up to period 15 (to May 8, 2021).

Version 3.0 of the database also includes aggregated values taken directly from the CRA website for the fifteen periods at the province/territory geography level and at the industry sector level (i.e., two-digit NAICS codes). The CRA website values are more current than the values for the periods published in previous versions (Version 1.0 and Version 2.0) of the database and account for late filers and revised claims. For any data point available from the CRA website, both the website value and the corresponding estimates developed by processing the CRA microdata file are provided.

4. Compilation methodology

This section provides an overview of the processing done to compile the database from the CRA source data files.

4.1 Data linkage

The first processing step consisted of linking the CEWS microdata to the Business Register in order to identify each business’ industry and geography classification. Each record in the CEWS microdata data file contained the unique business number (BN) associated with a business. Within the BR, BNs are mapped one-to-one with unique operating entity numbers (OENs).^Note  BNs from the CRA source data file were thus deterministically mapped using the BN-to-OEN link on the BR.

4.2 Aggregation and allocation of variables

In order to provide granular data at the census subdivision level and by NAICS, the CEWS source file required: (i) aggregation to the business enterprise level; and, for complex businesses with multiple locations, (ii) allocation to the business locations which are associated with the enterprise.

In cases where a business enterprise reported only one location, the allocation was straightforward. However, some records were for multi-location businesses, e.g., a head office with at least one other (branch) location. Multi-location businesses represented approximately one-twentieth of locations and one-third of the CEWS subsidy approved over the fifteen periods covered in this database. The allocation of CEWS data to individual locations was done in order to reflect the use of CEWS by multi-location businesses more accurately.

The Business Register’s Factor Allocation Table was used for the allocation process. The Factor Allocation Table provides two sets of factors: revenue factors and employment factors. They show, respectively, the share of the enterprise’s total revenue or employment belonging to each of its operating entities.^Note 

As an example, consider an enterprise in which revenues and expenses are reported from branches to a head office, which then reports for the entire business enterprise to the CRA. By using the Allocation Factors, it is possible to allocate the business enterprise aggregations, i.e., revenue and expenses from the head office, back down to the local branches. This enables the enterprise’s financial variables to be processed at a granular level geographically and by industry.

4.2.1 Allocation of CEWS claim amount and supported employees

For the database, preference was given to using employment allocation factors (the proportion of an enterprise’s employees belonging to each operating entity) when allocating aggregated multi-location business variables to each business location. Using employment allocation factors intuitively aligned with CEWS given that it is a program that targets employment support via a wage subsidy.

When employment allocation factors were not present in the BR’s Factor Allocation Table for a multi-location business, or when the sum of its employment allocation factors did not add up to 100%, then revenue allocation factors were used instead.^Note  This derivation was required in only a small number of cases.

4.2.2 Allocating the indicator of retroactive rehire

The indicator of retroactive rehire variable, like all other variables drawn from the CRA source data file, was only available at the payroll number level for each business. As a result, these values were also aggregated to the business enterprise level, before being allocated down to the business location level.

However, it was not possible to allocate this down to the business location level precisely, given that the allocation process using the Factor Allocation Table of the BR does not lend itself well for this purpose. Note that, in general, there is no one-to-one correspondence between payroll numbers and business locations (for complex businesses). As a result, this indicator only denotes that the business has rehired one or more employees who had been previously furloughed somewhere in their business structure. In the case of multi-location businesses, e.g., a business with a head office and at least one other separate branch location, this rehire flag would only be reported for whichever branch processed that payroll account. It would not specify, exactly, which business location(s) the retroactive rehire indicator is associated with; i.e., the exact business/branch location in which previously furloughed employees were rehired.

In order to allocate this indicator to the business location level, the following rule was applied: whenever the rehire flag was equal to 1 (i.e., indicating a rehire took place) in a given CEWS reporting period for any business enterprise associated with a multi-location business, e.g., a head office location, then all locations associated with that enterprise for that CEWS reporting period were also coded as rehire equal to 1.

4.3 Allocation of values to geographic units

The Standard Geographical Classification 2016 (SGC) was used for geographic allocation of CEWS variable values based on business location. SGC provides a systematic classification structure that categorizes all of the geographic areas of Canada.^Note  The SGC comprises a hierarchical coding system that provides a unique identifier for each level within the geographic hierarchy. The geographic levels used in the database are: Canada, provinces and territories, census metropolitan areas/census agglomerations (CMA/CAs), and census subdivisions (CSDs).

A rural and urban breakdown at the Canada and provincial/territorial levels was also created. CSDs outside CMA/CAs were classified as “rural”, while those inside CMA/CAs, as “urban”. Based on whether a business location’s CSD in the BR was rural or urban, it was aggregated into one of these categories.

The Business Register contains the location data associated with every business in Canada. Once the link between the CEWS source data file and the BR was established, a unique SGC was derived for each business location using the geo-coordinates of the business location recorded on the BR. In a few instances, no location data was present within the BR, or the location given was outside Canada. Manual search was performed to assign locations to these businesses. Businesses for which an address could not be found were dropped, and for businesses for which multiple locations were found, the values were allocated in equal parts to all locations (e.g., for a two location enterprise, each location was allocated 50% of CEWS subsidies).

4.4 North American Industry Classification System industry assignment

The North American Industry Classification System (NAICS) is used to classify businesses into industry sectors.^Note  The database provides data at both the 2- and 3-digit NAICS levels. At the 2-digit NAICS codes, industry sectors are defined, whereas at the 3-digit NAICS codes, industry subsectors are defined, whereas at the 3-digit NAICS codes, industry subsectors are defined.

NAICS codes were assigned to each business location as they were listed in the Business Register. If a NAICS code was not present for a business location in the most recent version of the BR at the time of processing, the previous BR was then examined to see if a NAICS was assigned for that business location. This process was done iteratively for previous BR versions until a NAICS code was found.

Nevertheless, it was not possible to find the NAICS codes for a small number of businesses, which were assigned the 2-digit NAICS code ‘99 - Unknown’. In addition, to reduce the potential suppressions for NAICS with small counts, two additional 2-digit NAICS code were merged with “99 – Unknown”. These are: NAICS 81 “Other Services (except Public Administration)”, and 91 “Public Administration”.

4.5 Protecting confidentiality

Given the high level of granularity provided by the database (geographic and industry), data suppression was implemented to preserve confidentiality using a hierarchy-aware confidentiality program developed by Statistics Canada (G-Confid).

In general, there are two reasons why a value may be suppressed to protect confidentiality:

1. Primary suppression
   1. There are few businesses contributing to the value of a cell; or,
   2. A cell value is dominated by one or a few businesses.
2. Secondary suppression: another cell is suppressed to protect the value of a primary suppressed cell so a higher-level aggregate can be published. For example, if the value of a variable for only one CSD in a particular province and industry is suppressed for confidentiality reasons, a second CSD will also have this value suppressed in order to prevent the suppressed value from being derived from the provincial total.

5. Database contents

As indicated earlier, the database is organized as follows:

1. Main variables: Start date of period, Total basic CEWS subsidy, Total number of CEWS supported employees, Number of business locations claiming CEWS, and Number of business locations indicating retroactive employee rehire;^Note 
2. Geography levels: Canada, province/territory, rural and urban parts of a province/territory, census metropolitan area (CMA), census agglomeration (CA), and census subdivision (CSD); and
3. Industry levels: industry sector (2-digit NAICS) and subsector (3-digit NAICS).

See the data dictionary in section 6 for details on the variables.

6. Data dictionary

This data dictionary below describes the variables of the database.

7. Possible future updates and revision strategy

As indicated above, by virtue of the fact that claimants are able to retroactively apply for CEWS (generally up to 180 days after the end of a claim period), the subsidy amounts and supported employees can continue to change as more applications are received and approved, or reassessed and not approved by the CRA.^Note  For this reason, future updates of the database are likely to result in revisions to earlier periods of data. Updated aggregate values as available on the CRA website will be included in a subsequent release, however, breakdowns available only in this database will not be revised. A future release covering more recent CEWS periods is planned, although no date has been determined as of the time of release of Version 3.0 of the database.

8. Downloading the Canada Emergency Wage Subsidy Regional and Community-level Database

For ease of download, the Canada Emergency Wage Subsidy Regional and Community-level Database is provided as a compressed comma-separated values (CSV) file.

• CEWS Version 3.0