The measures

Measure 1: Estimating environmental employment using aggregate-level data
Measure 2: Estimating environmental employment using firm-level ratios
Measure 3: Estimating environmental employment using past survey responses to questions on environmental employment
Measure 4: Estimating environmental employment using concordance list based on the Input-Output Classification

Measure 1: Estimating environmental employment using aggregate-level data

Given the absence of published environmental employment data based on direct survey results, most analysts use aggregate-level ratios in order to estimate the industry's environment-related employment. Aggregate-level ratios are computed using the data from published tables in the Environment Industry Survey report.² Tables 1 and 2 illustrate the calculation of environmental employment estimates using this method. In the two reference periods, 1998 and 2000, the share of environmental revenues remained virtually unchanged at the Canada level. However, the ratio of environmental revenues to total revenues at the industry group level showed more variability, hence resulting in a wider range of environmental employment estimates between 1998 and 2000.³

Throughout the report, this method will be used as the benchmark for comparison with other alternative approaches in estimating environmental employment.

Top of Page

Measure 2: Estimating environmental employment using firm-level ratios

One of the alternatives to aggregate-level ratios is the use of firm-level ratios. This approach aims to minimise the influence of the top contributors in the calculation of environmental-revenue-to-total-revenue ratios for each industry group. The following methodology is applied to all establishments in the EIS:

• For each establishment, calculate the ratio of environmental revenues to total revenues.
• Apply the respective ratios from #1 to the establishment's total employment, in order to obtain an estimate of environmental employment.
• Use the establishment-based results from #2 to produce an aggregate table by industry group.

Tables 3 and 4 present the results of this approach for 1998 and 2000. Figures 1 and 2 show that in both 1998 and 2000, Measure 2 (firm-level approach) yielded higher environmental-revenue-to-total-revenue ratios in the majority of industry groups compared with the ratios calculated using Measure 1 (aggregate-level approach). This observation was true in 14 out of the 29 industry groups in 1998; and in 20 out of the 29 industry groups in 2000. Upon examination of the resulting total environmental employment estimates, however, Measure 2 produced a lower estimate of 90 883 environmental employees than Measure 1 (95,041) in 1998. The reverse was true in 2000, where Measure 1 showed a drop in total environmental employment (87,492) and Measure 2 presented an increase (93,981). These differences illustrate how the environmental-revenue-to-total revenue ratios of the top contributors affect the overall estimates using the two approaches. Given the more detailed data used in the firm-level approach, one might lend greater credence to the results of Measure 2.

	Figure 1. Shares of environmental revenues to total revenues, by industry group, 1998: Comparison between aggregate-level and firm-level approaches
	Figure 2. Shares of environmental revenues to total revenues, by industry group, 2000: Comparison between aggregate-level and firm-level approaches

Graphical comparisons of environmental employment estimates using aggregate-level ratios and firm-level ratios are shown in Figures 3 and 4. In both years, environmental employment estimates in most of the industry groups were higher using firm-level ratios in comparison with the equivalent aggregate-level ratios. Looking at the key industry groups: Waste Management and Remediation Services (28), Engineering Services (19), Recyclable Materials Wholesalers (14) and Construction Engineering (4), the two approaches yielded very similar results, with the exception of Engineering Services (19).

	Figure 3. Estimates of environmental employment, by industry group, 1998: Comparison between aggregate-level and firm-level approaches
	Figure 4. Estimates of environmental employment, by industry group, 2000: Comparison between aggregate-level and firm-level approaches

In the case of Engineering Services (19), the relatively lower environmental employment estimates using firm-level ratios could be attributed to the nature of this industry's workforce and the diverse projects undertaken by engineering firms. For instance, some engineering companies had less than 1,000 employees but were involved in large, multi-million dollar environmental projects. In these instances, a smaller group of highly-skilled professionals could be involved in the administration of these projects, and the rest are contracted out to smaller engineering firms. On the other hand, there were also cases of other engineering firms that reported a large number of employees such as technicians, support staff, field workers, etc. who may have been involved in smaller, short-term environment-related projects. Neither one of the two approaches could be used to describe the composition of the personnel involved in the production of environment-related goods and services.

Top of Page

Measure 3: Estimating environmental employment using past survey responses to questions on environmental employment

In the Environment Industry Survey, 1996 and 1997, a section of the survey was dedicated to questions related to total employment and environmental employment, including specific employee categories. The section was scaled down in the Environment Industry Survey, 1998 due to low response rates to the environmental employment questions.

The proportion of environmental employment to total employment could also be estimated using the micro-level data for reference years 1996, 1997 and 1998. Under this approach, the data are based on actual survey results (i.e., responses to the environmental employment questions) for the three years.⁴

For data comparability, only the 745 establishments who responded consistently to the question on environmental employment in 1996, 1997 and 1998 were included in the table. In addition, the 1996 and 1997 survey results by industry group were based on the Standard Industrial Classification (SIC), therefore, establishments had to be classified to their corresponding industry groups based on the North American Industrial Classification System (NAICS) in order to produce a coherent set of estimates. Table 5 presents the environmental employment to total employment ratios by industry group. This approach produces relatively lower ratios, and therefore, lower environmental employment estimates in comparison with the first two approaches.

Further analysis was conducted by comparing the environmental revenue to total revenue ratios against the environmental employment to total employment ratios of the establishments in the sample. The rationale behind this comparison is to evaluate the hypothesis that environment-related activities are reflected in both the revenue and employment variables of firms and that the environmental revenue ratios should closely correspond to the environmental employment ratios for each industry group. Table 6 shows the average ratios by industry group for reference year 1996, 1997 and 1998. At the Canada level, there would be an 8% differential between the two methods of estimating for environmental employment. The use of average environmental employment ratios will produce higher environmental employment numbers in 16 out of the 29 industry groups. The data for the three of the top four industry groups seem to support the above hypothesis: Engineering Services (19), Recyclable Materials Wholesalers (14) and Construction Engineering (4).

Table 7 illustrates the three approaches (aggregate-level ratios, firm-level ratios and ratios based on direct survey responses) using data from the Environment Industry Survey, 1998. Industry group shares of total environment industry revenues as well as the response rates for key variables (environmental employment and environmental revenues) are also provided in Table 8 to help in the evaluation of environmental employment numbers generated by the different approaches.

Top of Page

Measure 4: Estimating environmental employment using concordance list based on the Input-Output Classification

A fourth method that could be used to produce environmental employment estimates is the use of the Input-Output (I/O) tables.⁵ The proposed methodology is as follows:

• Establish a concordance between the list of environmental goods and services from the Environment Industry Survey and the list of all commodities based on the Standard Classification of Goods (SCG).⁶
• Develop a second concordance list that matches the SCG-based list of environmental goods and services (from #1) and list of I/O commodities.
• Using the resulting list from #2, identify the I/O industries that have environment-related activities.
• Calculate the revenues shares of environment-related activities for each industry in #3 based on the results of the Environment Industry Survey.
• Assuming that environmental revenue shares could be used as proxies to estimate the proportion of total employees who worked directly or indirectly in the production of environmental goods or in the provision of environmental services, apply the shares from #4 to the overall industry employment⁷ in order to produce estimates of environmental employment by industry.

This limited research project estimates environmental employment for the non-surveyed portion of the environment industry⁸ and to further refine and validate results from the surveyed portion. The results of this preliminary research will be used to evaluate the possibility of developing a simulation method using the I/O framework to develop a more complete profile of Canada's environment industry.

The following tables have been finalised:

• Table 9: Environmental goods and services concordance between EIS categories and SCG codes
• Table 10: I/O industries with environment-related activities
• Table 11: Environment Industry Survey results and industry gross output, by I/O industry, 1998
• Table 12: Estimates of environmental employment using concordance list based on the Input-Output Classification (measure 4), by I/O Industry, 1998

Based on comments and recommendations from Industry Canada, additional tables by province and by establishment size were generated. The estimates for these additional tables were based on the type of Measure which Industry Canada deems to be most appropriate for producing estimates of environmental employment:

• Table 13: Estimates of environmental employment based on firm-level ratios (measure 2), by province and territory, 1998
• Table 14: Estimates of environmental employment based on firm-level ratios (measure 2), by province and territory, 2000
• Table 15: Estimates of environmental employment based on firm-level ratios (measure 2), by establishment size, Canada, 1998
• Table 16: Estimates of environmental employment based on firm-level ratios (measure 2), by establishment size, Canada, 2000

Next steps

This report presented tables by industry groups. The results of the Environment Industry Survey 2002 will be published in the Spring 2004. The report will include the survey response to the sub-component employment question.

In terms of the use of the Input-Output (I/O) tables in estimating environmental employment, this aspect of the project evaluated and established a concordance between the list of environmental goods and services from the Environment Industry Survey and the list of all commodities based on the Standard Classification of Goods (SCG). In addition, the project developed a second concordance list that matches the SCG-based list of environmental goods and services and list of I/O commodities and I/O industries. These lists could be used to identify the "environment-related" industries and/or commodities which would be critical to the development of an I/O model or framework specifically for the environment industry.