3-year Employment Outlooks - Methodology

The Employment and Social Development Canada's (ESDC) 3-year Employment Outlooks estimate job outlooks for the 516 occupations of the National Occupational Classification (NOC), by province, territory, and economic region, data permitting.

The 3-year Employment Outlooks are based on a model developed by ESDC and Service Canada regional economists, which includes the use of data sources available in all regions, proven statistical techniques, and significant analytic support from labour market specialists, who are active in each province. The objective of the methodology is to assess whether the employment outlook for a given occupation within a specific province or economic region, is very good, good, moderate, limited or very limited. ESDC is the only organization in Canada to produce outlooks for all occupations and regions, covering all Canada, on an annual basis. The forecast covers a period of 3 years because a longer outlook could potentially overlap two economic cycles, and would be too close to the 10-year forecast done by the Canadian Occupational Projection System (COPS) model.

Main Indicators

Four indicators serve as the foundation of the outlooks exercise. Each indicator is calculated for each of the 516 NOCs by province, territory, and economic region, data permitting.

Indicator 1: Employment Growth Rate

The first indicator considered when assessing employment outlooks is the annual employment growth rate by occupation over the 3-year forecast period.

First, an industrial employment forecast is developed at the provincial and economic region levels, based on a custom industrial structure developed to take into account the main sectors of the provincial or regional labour market.

Next, this industrial employment forecast is used to estimate occupational growth based on detailed information available regarding the occupational composition of labour within each industry.

Finally, these occupational projections are adjusted to take into account the impact of structural changes, such as the introduction of new technologies or new ways of organizing work, as well as situational factors related to economic cycles.

The statistical technique used to estimate these yearly matrices is called the Maximum Entropy, which provides a constructive criterion for setting up probability distributions (i.e. matrices) on the basis of partial knowledge. It allows estimating missing values when disaggregating data to achieve more granular information. The constraints used to perform this function are the known data at the aggregate level (employment by industry and employment by occupation from the Labour Force Survey) and the base coefficient matrix from the Census.

Lastly, a series of regressions are used to establish variable coefficients, which allow adjusting entropy estimates, to reflect long-term employment trends within each of the occupations.

Indicator 2: Replacement needs rate

The total number of projected retirements and deaths for the 3-year forecasting period is estimated.

This indicator is defined as the ratio of the estimated attrition (retirements and deaths) over the projection period, in relation to the total employment projected for the same period. The forecast of replacement needs by occupation is performed in two steps. First, a global estimate of replacement needs is performed for the province using demographic data. This global estimate is then distributed across occupations on the basis of attrition rates from the Canadian Occupational Projection System (COPS) model, available at the national level, and the age composition of each occupation from the Census.

Indicator 3: Experienced unemployed workers index

The number of experienced unemployed workers at the beginning of the forecast period, expressed as a percentage of employment during the same period and weighted by the relative length of the benefit period is estimated.

The indicator is determined using Employment Insurance administrative data collected by ESDC. A series of calculations are undertaken in order to obtain detailed estimates by economic regions and province. In particular, the process involves dividing the number of beneficiaries by the average three-year projected employment levels for each occupation.

Indicator 4: Net needs (Composite Indicator)

Net needs is a composite indicator based on the results of the three preceding indicators. It is calculated by summing, for each occupation, province, territory and economic region, the average employment growth (indicator 1) and replacement needs (indicator 2) during the forecast period to determine gross needs, and subtracting the average number of available experienced workers at the beginning of the period (indicator 3). For the latter, the average number of beneficiaries for the three most recent years is used. The sum is then divided by the average employment during the forecast period to express the indicator as a proportion.

The indicator corresponds to the total needs due to growth, retirement and deaths over the forecast period that cannot be satisfied by experienced unemployed workers available at the beginning of the period.


Determining the Employment Outlooks

As mentioned above, the Employment Outlooks seek to assess whether employment outlook for an occupation within a specific province, territory or economic region, is very good, good, moderate, limited or very limited. This is determined using a four-step approach:

Step 1 - Establish a Frame of Reference

Using Statistics Canada's demographic projections and the Labour Force Survey, a first analysis of the current and projected labour market situation is conducted. It includes a number of elements, such as

  1. A preliminary global (macroeconomic) scenario (e.g. COPS, consensus from financial institutions, etc.)
  2. Demographic projections (Statistics Canada, provincial governments, custom projections)
  3. Labour market indicators by age group (total population, labour force, employment, unemployment, participation rate, employment rate, unemployment rate)

Step 2 - Generate Mechanical Results

Once the four indicators are computed for each occupation, each is given a rank of 1 to 6.

This score is determined at the provincial level by assessing the data over a historical period of 10 years, where possible. These results are then placed in order from the best situation to the worst and divided into six equal groups, taking note of the threshold (lower-end) values for each group. A rank of 1 is associated with occupations in the group with the best situation and so on with 6 being associated to occupations in the group with the worst situation. The results for the forecast period are then compared to the historical thresholds and the corresponding scores are attributed. Occupations with insufficient data to calculate results for a given indicator are excluded from the ranking for that indicator.

Example: Scorecard for an occupation (fictitious)

Rank Final Rank
Employment Growth Rate 1 9
Replacement needs rate 1
Experienced unemployed workers index 4
Net needs 3

The scores of all indicators are summed (total of 4 to 24) and the final outlook is determined from the total. Finally, the outlooks status (very good, good, moderate, limited or very limited) is attributed by applying the following table:

Final Score Outlooks
4 to 7 Very good
8 to 11 Good
12 to 16 Moderate
17 to 20 Limited
21 to 24 Very limited

At this stage, occupations which are excluded from ranking in any indicator is given an ¿undetermined¿ rank.

Step 3 - Validate Mechanical Results

The mechanical results are also validated through an extensive consultation process. Final changes - if any - are justified and the rankings standardized to ensure that the adjustments do not create inconsistencies among the regions and the ranking established for the province, where appropriate. Consultations include:

  1. Service Canada's regional economists validate each of the outlooks throughout the process and, at the end, the outlook status;
  2. Feedback is sought from provincial and territorial counterparts; and
  3. Alternative quantitative and qualitative data sources (e.g., Survey of Employment, Payrolls and Hours, Job Vacancy and Wage Survey), administrative data (e.g., Employment Insurance, Job Bank data on job postings), labour market news, research reports by industry and labour organizations are consulted to validate results.

Step 4 - Develop Trends Statement

Once the mechanical results have been validated, the next step is to develop trends statements (short explicative text) that accompany the outlook status and publish them on Job Bank. These trends statements often include information related to the indicator scores as well as information gathered by the economists during the validation stage. Trends statements include contextual information on the factors behind the call, as well as key statistics associated with the occupation.

Date modified: