Findings

View the most recent version.

Archived Content

Information identified as archived is provided for reference, research or recordkeeping purposes. It is not subject to the Government of Canada Web Standards and has not been altered or updated since it was archived. Please "contact us" to request a format other than those available.

Archived

This page has been archived on the Web.

Skip to text

• Why are natural resource sector value chains important?
• Overall, there was employment growth for the forestry value chain but employment declined in primary production and processing in most resource sector value chains
• Employment in primary production and services and inputs became more geographically concentrated
• Sector intensity: an increase in rural non- metro-adjacent regions, a decline in predominantly urban regions
• Services to primary and wholesaling were increasingly clustered or located in core production regions
• Value chain distribution across regions and regional milieu: spatial patterns
• Conclusions
• Appendix tables

Text begins

Why are natural resource sector value chains important?

A "value chain" can be defined as "the full range of activities which are required to bring a product or service from conception, through the intermediary phases of production, delivery to final consumers and final disposal after use" (Kaplinsky 1999:121). These activities can include input and services to the primary producer, primary production, processing, handling, transportation, storage and retail, and service activities related to processing and marketing, including financial and insurance services, etc. (Porter 1985). In this analysis we consider only four aspects of the value chain – primary production, services to primary, wholesaling and first-stage processing (Box 2).

Natural resource sectors have long been exposed to globalization (falling price of transportation and communications) and fluctuations in world prices (Innis 1933, 1940, 1951). But the evolving global environment and emerging competitors in the world market poses continuous new challenges to traditional areas of production and processing.

Over the last decades, the economy of rural regions has become increasingly diversified and service-oriented. This transformation is largely associated with major changes in the predominantly rural and rural northern resource sectors (agriculture, forestry, fishing, mining and energy). There are two salient features of these changes which are relevant for rural development initiatives and which can be better articulated by looking at the components of the value chain. First, there has been a continuous shift of employment from primary production to processing and service activities; for instance, from farming to services to farmers and processing of agricultural products. Second, there has been a spatial reorganization of some of the activities in the value chains, between and within predominantly rural and predominantly urban areas. For example, some of the services to farmers that were located in small villages have been relocated into larger towns or cities.

Both these processes of change are ongoing. New technologies keep reducing the amount of labour per unit of output in primary production (and increasingly also in the service sector). Similarly, the process of spatial reorganization has gone through rapid developments at the national and global level. Notably, after decades where processing activities became increasingly concentrated in core predominantly urban areas, many countries are now experiencing a reverse core-periphery pattern in which manufacturing activities are relocating into predominantly rural regions (Baldwin et al. 2001).

The analysis of value chains, from a regional perspective, requires defining and understanding of the ways in which specific economic activities of a region are linked to the rest of the national and global economy. The nature of these linkages can determine, to a large extent, the distributional outcomes along the chain and the capacity of a region to upgrade and sustain its economic base (Kaplinsky and Morris 2001). Part of the success of disadvantaged areas in generating employment from their primary products lies in the ability of these regions to access and take advantage of the benefit of specific value chains (UNCTAD 2000). In contrast, being cut off from a specific value chain may have severe consequences for a region. Therefore, an analysis of the structure and nature of predominantly rural value chains can help recognize persistent, and in some cases widening, regional inequalities (Kaplinsky and Morris 2001).

An analysis of predominantly rural value chains, even beyond the natural resource sectors, can provide better insights on how the rural economy often contributes to major value chains which are largely perceived as predominantly urban. Feser and Isserman (2008) show, for the USA economy, that a significant proportion of employment in some national value chains that are generally perceived to be urban based are, in fact, rural based. They conclude: "rural counties play a significant role in numerous, diverse value chains. This statement holds true whether one focuses on the primary, distinguishing core industries of a value chain or its higher wage industries. It also holds true for rural counties whether they are within metropolitan areas or outside such integrated regional economies. Rural counties do not always play a secondary role in key US value chains, nor are they always the location of lower wage segments in key chains. Thus, understanding and supporting the competitiveness of US industry entails recognizing and supporting its rural-based component." (Feser and Isserman 2008: 107).

Finally, even though the rise of the knowledge-economy has overshadowed the role of the resource sector in the economy of many OECD countries, the growing global concerns on climate change, energy production and sustainability of food systems have focused the attention back on the role played by the resource sectors in addressing these issues. The stewardship of natural and environmental resources will remain a distinguishing feature of predominantly rural areas. The use of these resources will continue to represent a unique asset upon which predominantly rural development initiatives can be built.

Overall, there was employment growth for the forestry value chain but employment declined in primary production and processing in most resource sector value chains

In 2001, total employment in all resource sector value chains amounted to 2 million workers, representing 13% of Canada's total employment (Table 1). Over the 1991 to 2001 period, resource sector value chain employment remained at about 2 million workers whereas total employment in all sectors increased 10% over this period, from 14.2 million to 15.6 million workers. As a consequence, the share of workers employed in resource sector value chains declined from 14% in 1991 to 13% in 2001.

Table 1 Employment in natural resource sector value chains, Canada, 1991, 1996 and 2001

Part of the transformation within resource sector value chains is reflected in the employment shifts between components of the value chains. The main trend recorded during the 1990s was a decline in primary production employment in most value chains and a decline in processing employment in most value chains. This general decline took place in the context of an increase in services and/or wholesaling employment in most value chains. In the case of forestry, this increase in services and wholesaling led to employment growth for the value chain as a whole.

Employment in the primary production component of resource sector value chains declined between 1991 and 2001, with an overall loss of about 120,000 employees. This trend is most evident in mining, energy and agriculture (Figure 1). In absolute terms, the decline was largest for agriculture (with a drop of almost 80,000 workers in primary production). But in terms of the rate of change, the contraction was more pronounced in mining (a decline of 30% or 24,000 employees) and in the energy sector (a decline of 20% or about 14,000 employees). For the fishing industry, primary production employment levels were essentially stable between the beginning and the end of the decade. Part of this trend may be the result of new developments in the aquaculture sector (Statistics Canada 2008).

Similar to primary production, employment in first-stage processing for all resource sectors combined was lower at the end of the 1990s, compared to the beginning of the 1990s. There was an overall loss of almost 20,000 employees. However, employment in forestry and mining processing was higher in 2001, compared to 1991. For these two sectors, however, the increase between 1991 and 2001 represented a partial recovery from a sharper decline in the late 1980s. First-stage processing in the fisheries and energy value chains recorded the largest decline both in absolute and percent terms, with job losses in the order of 20,000 workers in each sector and a relative decline of 31% and 18%, respectively.

In contrast with the trends for the primary production and first-stage processing activities, employment in services to primary increased significantly from 1991 to 2001. For the five resource sectors combined, employment in the services to primary component of the value chain grew almost 26,000 workers. Changes were particularly large for the services component of the energy value chain (increasing 59% or about 19,000 employees) and for agriculture (with a 29% growth or 14,000 additional employees).

Similarly, employment in the wholesaling component for resource sector value chains as a whole was on the rise between 1991 and 2001. The wholesale component for all resource sectors value chains grew by almost 50,000 workers, but most of this was due to expansion in wholesaling in the forestry sector (almost 45,000 additional jobs) plus an employment increase of 13,000 in agriculture wholesaling. However, energy and mining wholesaling recorded an employment decline.

The net effect of these employment shifts was some growth from 1991 to 2001 in the forestry value chain, driven in particular by wholesale and processing expansion. For other value chains, the net effect of the changes over this period has been a declining employment trend, although the expansion in service and/or wholesaling has partially offset the decline in primary and processing employment.

The largest resource sector value chain, as defined in our study, is the agriculture value chain. The level of employment in 2001 was 739 thousand (Figure 1). This is equivalent to 5% of total employment (Table 1). This is followed by the forestry value chain with 494 thousand being employed (3.2% of total employment) and the mining value chain with 426 thousand being employed (2.7% of total employment).

Figure 1  Total employment in natural resource value chains, Canada, 1991 to 2001

As noted in Table 1, 13% of total employment in Canada was dedicated to a resource sector value chain in 2001. However, this share varies considerably by type of region. In 2001, employment in resource sector value chains represented 25% of total employment in rural non-metro-adjacent regions (Figure 2). This is a decline of 2 percentage points over the 1991 to 2001 period. This decline in share in rural non-metro-adjacent regions is due to the combination of a growth in total employment (up 4% from 1991 to 2001) and a decline in resource sector employment (down 5% from 1991 to 2001) (Appendix Table A.16).

Figure 2 In 2001, 25% of employment in rural non-metro-adjacent regions was in a resource sector value chain

The resource sector employment share in rural northern regions recorded a larger decline in employment share – a decline of 4 percentage points between 1991 and 2001. This is again due to a growth in total employment and a decline in employment in resource sector value chains.

Not surprisingly, the share of employment in each resource sector value chain differs by type of region. In each of rural metro-adjacent regions and rural non-metro-adjacent regions in 2001, nearly 10% of total employment was in the agriculture sector value chain (Figure 3 and Appendix Table A.17)). Within rural northern regions, 10% of total employment was in the forestry sector value chain.

Figure 3 In 2001, 10% of employment in rural northern regions was in the forestry sector value chain

When we just look at the employment within resource sector value chains, we see that 43% of 2001 resource sector value chain employment in rural northern regions was in the forestry sector value chain (Figure 4 and Appendix Table A.17). Within the resource sector value chains, forestry is relatively more important in rural northern and rural non-metro-adjacent regions. Agriculture is relatively more important in rural metro-adjacent regions. Mining is relatively important within resource sector value chain employment in intermediate regions. Energy has as a relatively constant share of resource sector value chain employment in each type of region.

Figure 4 In 2001, 43% of all resource sector employment in rural northern regions was in the forestry value chain

When we look at the contribution of each component of resource sector value chains to total employment, we see that in 2001, the processing component contributed 6.3% of total jobs in Canada followed by a contribution of 4.1% by primary production (Table 1).

Again, the size of this contribution differs by type of region. In rural non-metro-adjacent regions, 12% of total regional employment is employed in the primary production component of resource sector value chains (Figure 5 and Appendix Table A.17). Most of this employment is in primary agriculture (Table 1). In both rural non-metro-adjacent regions and rural northern regions, 9% of total employment is engaged in the processing component of resource sector value chains.

Figure 5 In 2001, 12% of employment in rural non-metro-adjacent regions was in the primary component of resource sector value chains

When we consider the contribution of each component within each value chain, the contributions differ by type of value chain. For example, within the agriculture value chain, employment in primary production represented 55% of employment in 2001 (Figure 6 and Appendix Table A.17). On the other hand, within the mining value chain, employment in first-stage processing represented 80% of total employment in the value chain.

Figure 6 Employment shares in major components of natural resource value chains, Canada, 1991 to 2001

Again, these within-chain employment shares differ by type of region. In predominantly rural regions, over 40% of employment within resource sector value chains is involved in primary production (Figure 7 and Appendix Table A.17). Over 25% are employed in the processing component. However, among resource sector workers in predominantly urban regions, over 60% are employed in the processing component.

Figure 7 In 2001, in each type of region, processing contributes over 25% of overall employment in resource sector value chains

As noted above (Table 1), the decline in share of total employment contributed by resource sector value chains was due to combination of:

1. overall growth in employment; and
2. overall decline in employment across all resource sector value chains.

This pattern is evident in each type of region (Figure 8 and Appendix Table A.16).

Figure 8 Employment in resource sector value chains declined in each type of region, 1991 to 2001

Employment in primary production and services and inputs became more geographically concentrated

Within Canadian resource sector value chains, employment in primary production and services to primary tended to become more spatially concentrated over the 1991 to 2001 period. In general, primary production employment declined relatively more in areas where it was less concentrated, while services and inputs employment grew relatively more in areas in which they were already more concentrated.

Not surprisingly, the primary production component of each value chain is highly geographically concentrated. In other words, relatively few jurisdictions account for most of the employment in primary production in each resource value chain. Perhaps obviously, primary production is linked to the location of the resource. The locational Gini coefficient (Box 4), which is one of the widely-used measures of concentration, reflects this pattern. Values of the Gini coefficient close to 1 indicate a high degree of concentration of employment within a few census divisions, while values close to zero indicate a more even distribution across census divisions. The locational Gini coefficient for primary production is relatively high for fishing, mining, agriculture and forestry (Appendix Table A.2).

Employment in services to primary also showed a relatively higher level of geographic concentration – particularly for the fisheries, energy, mining and forestry with the Gini coefficient ranging between 0.7 and 0.9. In most cases, the geographic concentration of services employment increased over the 1991 to 2001 period.

In contrast, processing and wholesale activities were more evenly distributed across space, with values of the locational Gini coefficient generally ranging from 0.15 to 0.58. The notable exception was employment in processing for fisheries which was as concentrated as primary production employment (Gini coefficient over 0.9).

Over the 1991 to 2001 period, the geographic concentration of employment in processing activities increased for agriculture, fisheries and energy, while it declined for forestry and mining. These latter two sectors also had an expansion of processing employment in absolute terms over this period.

In this section we have presented the overall level of geographic concentration. The next question we want to address is: what is the relative employment concentration of various value chain activities across the predominantly rural to predominantly urban gradient, and how did this concentration change during the 1990s?

Sector intensity: an increase in rural non- metro-adjacent regions, a decline in predominantly urban regions

Recall that employment declined in most components of resource sector value chains in most types of regions from 1991 to 2001 (Appendix Table A.16). Here, we consider changes in relative intensity of employment, using a location quotient (Box 4) as our indicator. The benchmark value for calculating the location quotient is the national intensity (or the national employment share). Thus, if the employment share in a region declines more than the national average, the location quotient (or relative "intensity" or relative "specialization") would fall. If employment share in a region declined less rapidly than the national average, then the location quotient for the given region would rise – indicating an increase in the relative intensity of employment in the given region, relative to the national average.

Between 1991 and 2001, the intensity of employment (i.e. the location quotient) for employment in resource sector value chains increased in predominantly rural regions (and particularly in rural non-metro-adjacent regions). This result for the change in the location quotient is due to a greater decline in the share of resource sector employment in predominantly urban regions compared to predominantly rural regions. Thus, at the end of the period, predominantly rural regions were slightly more intensive in resource sector employment, compared to Canada as a whole (Figure 9).

However, employment in resource sector value chains, as a percent of employment in each region, fell from 1991 to 2001 (Figure 2, above). Thus, from 1991 to 2001, the share of total employment contributed by the resource sectors declined in predominantly rural regions but it declined less than at the Canada level and hence the intensity relative to the national level went up (i.e. the LQ increased).

Figure 9 Relative to the Canadian average, the intensity of employment in resource sector value chains increased marginally in rural non-metro-adjacent regions from 1991 to 2001

With respect to the services component of the resource sector value chains as a whole, the location quotient showed an increase in rural metro-adjacent and rural non metro-adjacent areas (Appendix Table A.3). During the 1990s, the rural northern regions and the predominantly urban regions recorded a declining value of the location quotient in the services component for resource sectors as a whole. The relative measure of localization for intermediate regions remained more stable.

In the agriculture value chain, rural non-metro-adjacent regions had a relatively higher specialization (relative to the Canadian average) in all components of the value chain (Appendix Table A.4). Moreover, all components within rural non-metro-adjacent regions registered an increase in their location quotient from 1991 to 2001. During the 1990s, employment in agricultural services activities increased approximately 40% in rural metro-adjacent and rural non-metro-adjacent regions (approaching the level of 15,000 and 10,000 employees in 2001, for the two types of regions respectively) (Appendix Table A.16). A change of similar magnitude occurred in intermediate regions (where growth of 36% resulted in about 15,000 additional employees in agricultural services by 2001).

In contrast, rural northern and predominantly urban regions further de-specialized from agriculture. For example, from 1991 to 2001, employment in agricultural processing was reduced by half in rural northern regions (from approximately 1,000 employees to about 500).

In the fisheries value chain, rural metro-adjacent and rural non-metro-adjacent regions saw an overall increase in their location quotient from 1991 to 2001 (Appendix Table A.5). The location quotients in rural non-metro-adjacent regions increased for all components of the value chain, particularly employment in primary production and services. The rural northern regions maintained a relatively higher specialization (i.e. higher location quotient) in fishing activities although they saw a decrease in the location quotient, except in services activities. Predominantly urban and intermediate regions saw a decrease in their relative intensity of employment in the fishing sector (i.e. a declining location quotient).

The location quotient for employment in forestry's primary production increased in each type of predominantly rural region (Appendix Table A.6). In contrast, the relative intensity of forestry wholesale employment decreased in each type of predominantly rural region. This was due to a large increase of forestry wholesaling in intermediate and predominantly urban regions (75% and 66%, respectively, corresponding to an additional 11,000 and 25,000 employees, respectively) (Appendix Table A.16). Rural metro-adjacent regions also experienced an employment growth of 53% (corresponding to about 6,000 additional employees in wholesaling of forestry products). However, this growth in employment in the wholesale component of the forestry value chain was less than the growth for Canada as a whole which resulted in a decline in employment intensity (i.e. a decline in the location quotient) in rural metro-adjacent regions.

In the mining sector, each type of predominantly rural region saw increases in location quotients in the primary production component. Rural northern regions are relatively more intensive in employment providing services to the mining sector – and this relative intensity increased during the decade. Rural metro-adjacent regions also saw an overall increase in their location quotient in the mining sector (due to an increase in primary and in processing), while the rural northern regions showed a decrease (due to a decrease in processing).

Finally, each type of predominantly rural region reported an overall increase in intensity of employment (a higher location quotient) in the energy value chain. The only decline in a component was in the services to energy in rural northern regions (Appendix Table A.9).

Services to primary and wholesaling were increasingly clustered or located in core production regions

Most of the research on resource-reliant communities has focused on the concept of local reliance on the sector (Natural Resources Canada 2001; White and Watson 2001, Stedman et al. 2007). We extend this concept in two directions; first, by looking at regional reliance in the context of its regional milieu; and second, by looking at the relationship of the region and the milieu of the region for different components of the same value chain.

We do this by considering jointly the indicators of intensity (location quotient) for a given region and the corresponding spatially lagged values for this region. A spatially lagged location quotient is simply the average value of the location quotient in surrounding regions. We use the Moran's I statistic and bivariate regressions between the location quotient for a given region and the average location quotient of surrounding regions (i.e. the spatially lagged location quotient) to investigate these patterns (Box 3). The location quotient for a particular region does not tell us about the milieu in which the region is located. In this sense, the location quotient for a particular region provides only a partial picture of the spatial intensity of employment along resource sector value chains. Thus, the information for any given region does not indicate the distribution of these units across space.

In simple terms, we address the following types of questions: Are the areas with a high intensity of employment in a given sector clustered together? Or are they randomly spread over space? Are the localities with high intensity of processing activities in a regional milieu characterized by a high intensity of primary production? The Appendix Tables A.9 to A.13 summarize the results of this analysis.

In the agriculture value chain, regions (census divisions) with a higher intensity of employment in primary agriculture tend to be clustered together (coefficients of 0.53 in 1991 and 0.49 in 2001, Appendix Table A.9); in other words, these regions tend to be situated in a primary agriculture milieu. In these spatial clusters, there is also a higher intensity of employment in services and wholesaling. In particular, services to agriculture appear to be strongly connected to a primary agriculture milieu (coefficients of 0.91 in 1991 and 1.03 in 2001). In the case of the agriculture value chain, the wholesaling activities included in this analysis relate to the wholesaling of inputs for purchase by farmers (Appendix Table A.1). Thus, it is not surprising that regions that have a higher intensity of agricultural wholesaling employment tend to be surrounded by regions with a higher intensity of primary agricultural employment (coefficients of 0.86 in 1991 and 0.67 in 2001).

For agricultural primary production, services, and wholesaling, the same type of spatial association runs in the other direction, although the magnitude of the coefficients varies. For instance, regions with a higher intensity of primary agriculture employment are surrounded by regions with a higher intensity of agriculture wholesaling employment (coefficients of 0.32 in 1991 and 0.40 in 2001). Similarly, agricultural service and wholesaling employment tend to be spatially clustered; for instance, the regression coefficient of intensity of regional agricultural services and the agricultural wholesaling milieu is 0.62 in 1991 and 0.89 in 2001.

In contrast, there is limited spatial association between first-stage processing of agricultural products and agricultural primary production, services and wholesaling; thus agriculture shows a rather clear level of spatial disjuncture at this stage of the value chain. Employment in food processing is not in the milieu of production, services to production or wholesaling (coefficients are small or not statistically significantly different from zero). This is due, in part, to the fact that there are only a few (relatively large) food processing facilities and often these facilities are located in a (more) urban location.

Regarding the fisheries value chain, primary production, service to primary production and fish processing all tend to be in a primary production milieu (coefficients for 1991 and 2001 range from 0.29 to 0.39, Appendix Table A.10) – although the relationships are not as strong as for agriculture. The indication that emerges from the coefficients of spatial association is that of geographic clusters of regions in which primary production, services and processing are located.

For the forestry value chain, the greatest degree of spatial association is found between primary production and services to primary. Regions with a higher intensity of primary forestry employment tend to be clustered in primary forestry milieus (coefficients of 0.38 in 1991 and 0.35 in 2001, Appendix Table A.11) and services to primary have a stronger connection to a primary forestry milieu (coefficients of 0.54 in 1991 and 0.42 in 2001). Furthermore, regions with a higher employment intensity in forest processing tend to be surrounded by regions with a higher intensity of primary forestry employment (coefficients of 0.55 in 1991 and 0.44 in 2001), although this degree of spatial association is not as strong in the opposite direction.

In the forestry value chain, the wholesale activities included in this report relate to the wholesaling of wood and paper products to consumers. Thus, wholesaling to consumers show a relatively modest degree of spatial clustering (coefficients 0.13 in 1991 and 0.24 in 2001), and more interestingly is negatively associated with the primary forestry milieu as forestry products are wholesaled to an urban market and forestry production is a hinterland activity.

A review of the mining value chain indicates that mines and their associated services and processing generally show a modest degree of spatial association, suggesting that these activities exist within the same census division (i.e. within the given region), as opposed to a cluster of CDs. There is no apparent "milieu" or group of census divisions that are relatively intensive in mining. More interestingly, there is a negative spatial association of processing with production, meaning that regions with higher intensity of employment in primary production tend to be surrounded by regions with lower intensity of employment in processing (coefficients of -0.34 in 1991 and -0.36 in 2001, Appendix Table A. 12), and vice versa. This is another way of saying that minerals are processed where they are mined.

Finally, regarding the energy value chain, regions with a higher intensity of primary production employment show some degree of spatial clustering among themselves (coefficients of 0.23 in 1991 and 0.28 in 2001, Appendix Table A.13). Services to primary production have a stronger spatial association with a milieu of primary production (coefficients of 0.42 in 1991 and 0.48 in 2001); even more so, wholesaling of primary production, which in the energy value chain includes pipeline transportation industries (Appendix Table A.1) is strongly associated with primary production (coefficients of 1.39 in 1991 and 1.17 in 2001). In contrast, processing employment shows very little evidence of spatial clustering with any other segment of the energy value chain or with other processing intensive CDs (coefficients are, in most cases, not statistically significantly different from zero).

Value chain distribution across regions and regional milieu: spatial patterns

As a final step of this analysis, we use the relationship between the location quotient for a specific area and the location quotient of the regional milieu (spatial lag) to identify;

1. areas that are relatively directly reliant on a resource value chain; and to identify
2. areas with an economy that is not directly related to a resource sector value chain but which is located within the milieu of a resource value chain and thus may be affected by the resource sector's economic trends due to geographic proximity.

The analysis focuses on the data for 2001; we show a set of maps for the components of the value chains that appear to be spatially correlated, as determined from the results presented in Appendix Tables A.9 to A.13. For instance, based on the indications emerging from Appendix Table A.9, we break the agriculture value chain into a group composed of primary, services and wholesaling, on the one hand, and processing, on the other hand. Hence, for the purpose of mapping, we recomputed the location quotients and their spatial lag for these re-defined components of each value chain.

The methods used for this purpose are presented in Box 3. In brief, each census division is assigned to one of four groups, depending on the combination of regional reliance (the location quotient of the census division) and the regional milieu reliance (spatial lag value of the same location quotient). The resulting groups are:

1. higher regional and higher regional milieu values (dark blue on the maps);
2. higher regional and lower regional milieu values (light blue);
3. lower regional and higher regional milieu values (light red); and
4. lower regional and lower regional milieu values (dark red).

The first group (dark blue) can be considered to be the "core reliant regions"; this identifies the census divisions with a relatively higher intensity of employment in that natural resource value chain and that are also located in a regional milieu with a relatively higher share of employment in that value chain. At the opposite end, the "non-reliant regions" are those regions that have lower than average employment in that value chain and are located in a regional milieu which also has lower than average employment in that value chain (dark red on the map).

The distribution of all regions (census divisions) across this classification is shown in Appendix Table A.14 while the distribution of only predominantly rural regions is shown in Appendix Table A.15. As noted, the maps were generated from groupings of selected components of some value chains in cases where the geographic pattern of the components was highly correlated.

The maps and the tables in the appendix illustrate two relevant aspects of the spatial distribution of resource value chain employment. First, the maps delineate the regional clusters that have a relatively stronger association with the value chain of natural resources. In particular, the dark blue areas are the core reliant regions, i.e. regions with relatively higher value of the location quotient surrounded by other regions with relatively higher value of the location quotient.

The appendix tables indicate that the percent of regions that fall into the core reliance group has increased from 1991 to 2001 for all natural resource sectors except the fishing industry. Employment growth in Canada has generally been in the non-resource sectors. Thus, at the Canada level, the share of employment in resource sectors has been declining. The increase in the number of regions with "relatively" higher employment in the resources sectors is due, at least in part, to the lower share of natural resource employment at the Canada level – which is the benchmark for calculating the location quotients in this study.

Second, these maps show that, although many regions are not directly reliant on natural resource value chains, their economy might be strongly affected by the performance of these value chains.¹ The most intuitive example is that of a census division that is not reliant on agricultural related employment but that is located in a milieu of agricultural regions. The use of indicators for regions and for the regional milieu (i.e. the neighbouring regions) allows us to assess the potential relevance that core economic activities may have when the regional milieu is considered.

In 2001, almost 20% of all regions were classified as agriculture core reliant regions (Appendix Table A.14). Another 12% of the census divisions were highly reliant on agriculture but were located in a regional milieu with low dependence on this industry. In contrast, in 11% of the regions, their economy was not directly dependent on agriculture but they were located in a regional milieu with a higher degree of reliance on agriculture. Although the identification of these regions is strictly based on a criterion of geographic proximity, their economic performance might be significantly affected by initiatives targeted to the agriculture value chain as their economy may be more closely linked to that of core reliant regions as compared to the remaining 57% of census divisions which have a lower reliance on agriculture and are located in a milieu that has a lower reliance on agriculture.

Map 1 shows the geographic pattern of regions by intensity of employment in agriculture primary production, services and wholesaling, while Map 2 shows the distribution of regions by employment intensity of processing activities within the agriculture value chain. Not surprisingly, employment in processing activities is more intensive in more urbanized areas, like southern Ontario and southern Quebec.

For the fishing industry in 2001, almost 15% of Canadian regions were core fishery reliant regions (dark blue), while another 9% of the regions were surrounded by a regional milieu with an above average reliance on the fishing industries' value chain (light red). Not surprisingly, the regions with a core reliance on the fishing value chain are limited to coastal locations in the Atlantic Provinces and British Columbia (Map 3).

Maps 4, 5 and 6 show the spatial pattern for the intensity of employment in the forestry value chain. Map 4 combines primary production and services, while the other two maps show wholesaling (Map 5) and processing employment (Map 6) (See also Appendix Table A.11). About one-quarter of Canadian census divisions (Appendix Table A.14) have a relatively higher intensity of employment in forestry and related services and they are located in a milieu of similarly intensive census divisions (dark blue in Map 4). In addition, approximately 15% of Canadian census divisions exist in a regional milieu with a relatively high intensity of forestry employment in the primary and services components (light red in map 4) (Appendix Table A.14).

Map 5 shows where wholesaling of forest products to consumers is relatively intensive. Most intensive regions are regions with cities and/or growing populations. These regions operate in a different spatial context than the forestry workers.

Map 6 shows where the location of a higher employment intensity in the processing of wood (e.g. sawmills and pulp and paper mills). These regions are essentially the same regions as where the forestry workers are located (Map 4). As indicated in Appendix Table A.11, first-stage forestry processing tends to be located in the regional milieu of primary production.

Employment intensity in the mining value chain is mapped in two main parts. Map 7 shows the spatial distribution of census divisions according to their employment intensity in primary production and services. Map 8 shows the spatial pattern of intensity of employment in wholesaling and processing. The maps show the highly specific nature of this type of production; only about 10% of regions are core reliant regions (Appendix Table A.15). Northern Ontario and some regions in Saskatchewan and Manitoba account for most of them. In contrast, wholesaling and processing are largely concentrated in southern Ontario and southern Quebec (Map 8).

Finally, Map 9 and Map 10 display the spatial distribution of the components of the energy value chain. Map 9 shows the distribution of regions for the combined value chain components of primary production, services and wholesaling employment. Two major regional clusters are evident in Alberta and part of Saskatchewan. Similar to employment in mining, only about 10% of regions are in core reliant regions for production and services, and even less (about 2%) are high reliant regions located in a non-reliant regional milieu (Appendix Table A.14).

The spatial distribution of processing employment in the energy value chain (refined petroleum products, other petroleum and coal industries, and electric power system industries) is shown on Map 10. About 20% of the regions are core reliant regions, which appear relatively evenly distributed across Canada. Part of the spatial pattern is due to the dispersed nature of electricity generation via hydro, coal or gas fired generation plants and nuclear generation plants.

Map 1  Employment in primary production, services and wholesaling within the agriculture value chain: pattern of regional intensity and intensity within the regional milieu, 2001
Map 2  Employment in food processing within the agriculture value chain: pattern of regional intensity and intensity within the regional milieu, 2001
Map 3  Employment within the fishing value chain: pattern of regional intensity and intensity within the regional milieu, 2001
Map 4 Employment in primary production and services within the forestry value chain: pattern of regional intensity and intensity within the regional milieu, 2001
Map 5 Employment in wholesaling within the forestry value chain: pattern of regional intensity and intensity within the regional milieu, 2001
Map 6  Employment in wood processing within the forestry value chain: pattern of regional intensity and intensity within the regional milieu, 2001
Map 7  Employment in primary production and services within the mining value chain: pattern of regional intensity and intensity within the regional milieu, 2001
Map 8  Employment in metal processing and wholesaling within the mining value chain: pattern of regional intensity and intensity within the regional milieu, 2001
Map 9  Employment in primary production, services and wholesaling within the energy value chain: pattern of regional intensity and intensity within the regional milieu, 2001
Map 10 Employment in processing within the energy value chain: pattern of regional intensity and intensity within the regional milieu, 2001

Conclusions

A value chain perspective can substantially enhance our understanding of the rural economy and the challenges faced by rural regions. A value chain analysis focuses on the way in which economic actors are linked to the broader economic context. The nature of these linkages will determine, to a large extent, the distributional outcomes of changes occurring in increasingly global production systems. Almost all products from Canada's resource sectors are priced in international markets and/or are sold into international markets. An analysis of the value chains that link rural economies to the rest of the national and international economy is of major importance to rural development initiatives.

In this bulletin we have outlined a definition of natural resource sector value chains by classifying relevant industrial sectors to each value chain. The analysis extends only to 2001 because the 2006 data were not coded to the same Standard Industrial Classification used in earlier census periods. The North American Industry Classification System was used in 2006 and it only extends back to 2001. We limit our focus to employment statistics and look at each component of each resource sector value chain to portray the relative intensity of employment in each region. Furthermore, we focus on the relationship between a region and its regional milieu. It should be emphasized that this analysis remains a first exploration of predominantly rural value chains which has focused on a single indicator, namely employment data. It should also be recognized that any standard industry classification system does not always fully capture and overlap with the concept of a value chain.

Our analysis shows that the relative intensity of employment in resource sector value chains increased slightly during the 1990s in the rural economy, relative to the nation as a whole. Although the rural economy is diversifying, the intermediate and predominantly urban regions are doing so at a faster pace.

The findings of this research show that the resource sector value chains are upgrading, as measured by a shift of employment from primary production to processing and services activities, and this shift has been more intense in rural metro-adjacent regions. Rural northern regions have experienced a relative decline in the intensity of their employment in the processing and services activities in resource sector value chains. This trend may increase their exposure to global competition in raw commodity production.

The analysis of the characteristics of a given region and a consideration of the milieu in which the region is located identifies also those regions that, although not directly reliant on a natural resource value chain, could be strongly affected by the economic performance of these value chains because the region is situated in a natural resource milieu.

Appendix tables

Appendix Table A.1  Composition of major resource sector value chains: Standard Industrial Classification codes
Appendix Table A.2  Locational Gini Coefficients
Appendix Table A.3  Location quotients for all resource sector value chains, Canada, 1991 and 2001
Appendix Table A.4  Location quotients for the agriculture value chain, Canada, 1991 and 2001
Appendix Table A.5  Location quotients for the fisheries value chain, Canada, 1991 and 2001
Appendix Table A.6  Location quotients for the forestry value chain, Canada, 1991 and 2001
Appendix Table A.7 Location quotients for the mining value chain, Canada, 1991 and 2001
Appendix Table A.8 Location quotients for the energy value chain, Canada, 1991 and 2001
Appendix Table A.9  The agriculture value chain: spatial association between the intensity of employment in a given region and the intensity of employment in the region's milieu, Canada, 1991 and 2001
Appendix Table A.10  The fisheries value chain: spatial association between the intensity of employment in a given region and the intensity of employment in the region's milieu, Canada, 1991 and 2001
Appendix Table A.11  The forestry value chain: spatial association between the intensity of employment in a given region and the intensity of employment in the region's milieu, Canada, 1991 and 2001
Appendix Table A.12  The mining value chain: spatial association between the intensity of employment in a given region and the intensity of employment in the region's milieu, Canada, 1991 and 2001
Appendix Table A.13  The energy value chain: spatial association between the intensity of employment in a given region and the intensity of employment in the region's milieu, Canada, 1991 and 2001
Appendix Table A.14   Distribution of all census divisions by regional and regional millieu intensity of the resource value chain, Canada 1991 and 2001
Appendix Table A.15   Distribution of predominantly rural census divisions by regional and regional millieu intensity of the resource value chain, Canada 1991 and 2001
Appendix Table A.16 Number employed in each natural resource value chain by type of region, Canada, 1991 and 2001
Appendix Table A.17 Percent distribution of employment with each type of region, Canada, 1991 and 2001

Note

1. It should be emphasized that the criteria for interaction is geographic proximity