INTRODUCTION
Starting with the Canada 235 Industry set, occupation data is available for the Canada Product. The occupation data include occupational employment, wages, hours, and core competencies for more than 500 distinct occupations.
Driving these estimates are national-level ratios of Employment by Industry and Occupation. IMPLAN applies these ratios to Employment by Industry Region Data and Impact Results to generate estimates of occupational employment, wages, hours, and the core competencies associated with those occupations.
This article describes the data sources and methods used to estimate the occupation ratios and core competency data for the Canada Product. For more information about using Canada Occupation data in an analysis, please see Using Canada Occupation Data.
DATA OVERVIEW
Canada Occupation provides occupation data in accordance with the National Occupational Classification (NOC) system. The NOC system classifies Canadian occupations hierarchically according to their broad occupational category, primarily based on the type of work performed or the industry of employment, and their Training, Education, Experience, and Responsibilities (TEER) category. The first digit of the NOC code represents the broad occupational category. For example, Sales and Service Occupations start with the digit 6. The second digit represents the TEER category and corresponds to the type and/or amount of training, education, experience and responsibility typically required to work in an occupation. Each additional digit, up to the 5-digit Unit Group NOC Code, adds more detail to specify the nature and characteristics associated with a given occupation. The most detailed NOC code is the 5-digit Unit Group NOC Code, of which there are 516 unique occupations.
DATA SOURCES
There are four main data sources for the Canada Occupation data. Each raw data source has various levels of industry coverage, geographic coverage, and occupation detail.
- Statistics Canada Census of Population
- Provides detailed data for the construction of national industry by occupation matrix
- Raw data released every 5 years in accordance with the Census of Population
- Industries are classified by various levels of NAICS codes, up to 4-digit NAICS
- Occupations are classified by various levels of NOC codes, up to 5-digit NOC codes
- Statistics Canada Provincial and Territorial Economic Accounts
- Provides hours worked data by industry and province/territory
- Industries are classified by various levels of NAICS codes, up to 5-digit NAICS codes in some cases
- Statistics Canada Labour Force Survey
- Provides data on hours worked by occupation for the development of an occupational adjustment factor
- Occupations are classified by 2- digit NOC codes
- Employment and Social Development Canada and Statistics Canada OaSIS Database
- Provides data on knowledge, skills, abilities, personal attributes, work activities, and work context by occupation
- New database versions are released whenever updates are made, typically every 5 years
- Occupations are classified by 7-digit occupational profiles, where the first 5-digits correspond to the NOC Unit Group and the last 2-digits reflect additional information provided as part of the OaSIS profile
Methods
EMPLOYMENT AND WAGE RATIOS
IMPLAN relies most heavily on the Census of Population data for the construction of the national industry by occupation matrix used for employment and wages. The data come from a custom-tabulated dataset from Statistics Canada that provides national-level data on the count of respondents with wages, salaries, and commissions by industry (classified by 4-digit NAICS codes) and occupation (classified by 5-digit NOC codes). It also provides the median wage, salary, and commissions for those respondents. IMPLAN utilizes this detailed data to develop industry-occupation ratios for both employment and wages.
A limitation of the Census of Population is that it employs a method called random rounding, wherein values are rounded to base 5 and can either be rounded up or down, in a random fashion. This results in inconsistencies among the hierarchical occupation levels (e.g. the sum of the number of occupations in the ten Broad occupation categories for a given industry does not equal the total number of occupations in that same industry). Therefore, when developing the occupation-industry matrix, IMPLAN re-develops new count estimates for each occupation level, which have been controlled to the ultimate control total, the total count of persons with wages for all occupations in that industry. In other words, in order to ensure consistency among the hierarchical occupation levels, all values have been re-estimated and controlled to the total number of occupations in that industry. This is done in a top-down approach.
An additional challenge of using raw data produced in this way is that values between zero and 5 can either be rounded up to 5 or rounded down to zero, such that there is no way to distinguish between a true zero and a false zero (e.g. it is zero because it has been rounded down). The data production employs the following process:
In cases where the occupation parent count is non-zero and there is at least one non-zero child occupation count, IMPLAN assumes that any child occupation zero is a true zero and distributes the new estimated parent value to its children based on the shares in the raw data. If, however, the parent occupation is non-zero and all of the children occupation counts are also reported as zero, we must estimate values for the children such that the sum of the children is equal to the parent. In this case, IMPLAN estimates occupation employment distributions from higher-level NAICS codes. For example, IMPLAN’s Stone mining and quarrying industry maps to the 4-digit NAICS code 2123. Within the raw data, NAICS code 2123 has a value of 20 people employed in the 3-digit NOC code 321 Technical occupations in health (except practitioners of natural healing). However, all three of its children (3210 Technical occupations in therapy and assessment, 3211 Technical occupations in dental health care, and 3212 Medical technologists and technicians) have zeroes. Therefore, we look to the parent NAICS code (3-digit NAICS code 212) to inform the distribution from the parent NOC 321 to its children. This process is utilized more prevalently in unique or unlikely industry-occupation pairs (e.g. healthcare occupations in mining) and at more detailed NOC codes. For this reason, some industry-occupation pairs may be reflected in the IMPLAN data that are not reflected in the raw data.
HOURS WORKED
IMPLAN Canada Occupation hours data come primarily from Statistics Canada’s Provincial and Territorial Economic Accounts. This dataset provides data on annual average hours worked by industry and by province/territory. IMPLAN uses this data as the foundation for hours worked but incorporates occupational detail from the Labour Force Survey by calculating and applying an occupational adjustment factor. The occupational adjustment factor is calculated using the average hours worked for a given occupation divided by the average hours worked for all occupations. This assumes, for example, that a high-level manager works 25% more hours than the average occupation regardless of the industry employed.
CORE COMPETENCY
Core competency data comes from the OaSIS database. The OaSIS database expands the 516 NOC 5-digit Unit Groups to more than 900 7-digit occupational profiles. The 7-digit occupational profiles are built upon the NOC classification, where the first 5-digits correspond to the NOC Unit Group and the last 2-digits reflect additional information provided as part of the OaSIS profile. IMPLAN derives core competency data for each occupation by aggregating the more detailed occupational profiles to the 5-digit Unit Group NOC level.
RELATED ARTICLES
Canada Occupation Data Release Notes (Industry 236)