0354-A2

Forest reliance and community well-being in rural Canada

Richard C. Stedman, John R. Parkins and Thomas M. Beckley ¹

Abstract

The forest products sector is a major employer in much of rural Canada, and it is often assumed by policy makers that increased timber harvest is a viable means of rural economic development. Despite burgeoning research in the United States of America, relatively little attention has focused of the relationship between forest dependence and well-being in rural Canada, especially large-scale studies that compare across regions and/or forest sectors. This paper presents an overview of the relationship between forest dependence and well being in Canada. Analysis of 1996 Statistics Canada data revealed a great deal of variation in the effect of forest dependence on indicators of well-being (e.g. human capital, unemployment, income): some sectors had fairly positive outcomes (e.g. pulp and paper), others had more negative outcomes (e.g. logging). These relationships, however, vary a great deal by region, suggesting the need for more mid-range models that incorporate the particulars of place and sector.

Introduction

The relationship between the forest industry and the well being of people that depend on it for their economic livelihoods is of strong interest in Canadian forest policy. Canada has approximately 418,000,000 hectares of forest, with 119,000,000 hectares managed primarily for timber production, and another 249,000,000 hectares where forest harvesting is allowed (CCFM 1998). Canada is a world leader in producing wood and paper; 1996 exports were valued at more than 32 billion dollars (nearly equivalent to energy, fishing, mining, and agriculture combined). More than 825,000 Canadians work in forestry, and the sustainability of the human communities involved is an important goal of forest production (CCFM 1998). Research has begun to examine the relationship between community dependence on forestry (and other resource industries) and indicators of community well being (Parkins et al. 2001; Stedman 1999; Parkins and Beckley 2001; Beckley et al 2002). These studies usually focus at the community level, rather than making broad, data-based comparisons across communities and resource industries.

This question is of interest in other countries as well. Research in the United States has found negative outcomes of forest reliance on economic indicators of well-being, such as community stability (Kaufmann and Kauffman 1946), poverty rates (Bliss et al. 1992), and unemployment (e.g., Howze et al. 1993), and higher rates of social pathology such as divorce (Drielsma 1984) or crime (Force et al. 1993). Given the attention put on forest harvesting and wood processing as an economic development strategy in Canada, these findings should raise concerns in Canada, as well as other nations that rely on wood products industries for economic development.

The relationship between forest reliance and community well being may vary according to the particulars of region and the particular sector on which communities depend. A key question that has begun to receive attention is the degree to which this relationship varies between forest industry sectors, such as pulp and paper or logging (Overdevest and Green 1996) and between different locales or regions (Fisher 2001). Randall and Ironside (1996) make the point that too much emphasis has been placed on looking at the similarities between resource dependent places, with less focus on differences between sectors and regions. Our paper examines variation in the relationship between resource reliance and well being, by forest sector, region, and indicator.

Measuring Forest Reliance and Well Being

The data in this paper are from the 1996 Census of Canada (Statistics Canada 1998). In total there are 5,243 Census Subdivisions (CSDs) for 10 provinces and two territories. These data are collected every 5 years. Community reliance on natural resource industries (agriculture, forestry, fisheries, mining, and energy) can be measured several ways, but for the purpose of comparing to work in the U.S., this paper measures reliance as the proportion of employment in forest industries (including logging, sawmills, pulp and paper, and other value-added services such as reforestation and fire services).

Although much research in the United States defines well-being using poverty measures, other indicators of well-being are also important. Consistent with past research that suggests a broadened suite of indicators (e.g., Beckley and Burkosky 1999; Stedman 1999), this paper utilizes data readily available from Statistics Canada on CSD-level rates of family poverty, individual unemployment, median family income, and individual educational attainment.

Results

How Forest-Reliant is Rural Canada?

First, we describe the degree of forest reliance in Canadian rural CSDs. Because most resource extraction takes place in rural areas, urban CSDs (n=588) are excluded from the analysis. As well, there are 793 CSDs where data are suppressed. Generally, these involve First Nations native reserves where there are concerns about data quality and/or CSDs with extremely small populations where issues of confidentiality result in data suppression. Once these are excluded, 3,856 CSDs remain for analysis at the national level.

Canada appears to be more forest dependent than the United States. In the United States, 20% employment has been used as the cutoff for high levels of dependence (e.g., Elo and Beale 1985). However, many studies use a 10 percent criterion because the 20% criterion results in few cases. In contrast, nearly one-fourth (23.8%) of all rural CSDs in Canada (918 of 3853) have at least 10% of employment in forest related industries; 10.0% (386) have at least 20% of their employment in forest industries. This varies by region: table 1 illustrates the mean proportion of employment in the forest industry (disaggregated by sector) by region.

Table 1. CSD-Level Forestry Employment

¹ Atlantic = Newfoundland, Prince Edward Island, Nova Scotia, New Brunswick; Central = Quebec, Ontario; Prairie = Manitoba, Saskatchewan, Alberta; BC = British Columbia; North = Yukon, Northwest Territories. Provinces were aggregated into regions based on spatial proximity and similarity in resource dependence. By the latter criterion, British Columbia constitutes its own region based on its dissimilarity in forest dependence to any other province.

Nationally, 6.5% of employment in rural Canadian CSDs is in the forestry industry. These figures vary by region: rural British Columbia is dominated by the forest sector (16% of employment). Both Atlantic and Central Canada also have a strong forest industry presence, with over 5% of employment in the forest industry (the prairie region tends to be dependent on agriculture, and much of Northern Canada lies north of the treeline).

Forest Reliance and Community Well-Being

Overall, forest dependence appears to be associated with negative outcomes for rural Canada: higher rates of forest reliance are associated with lower educational attainment, higher rates of family poverty and unemployment, lower median family income, and 5 year migration rates (table 2). The relationship between forest reliance and unemployment is especially strong (r=.228); the relationship with median family income is modest (r= -.056). How do these figures vary by region (table 2).

Table 2. Regional Variation in Forest Reliance and Well Being.

Table two demonstrates much regional and indicator-based variation in the relationship between forest reliance and well being. Forest reliance appears to have particularly negative consequences in Central Canada, with stronger negative effects than the national average on educational attainment, migration, and median family income. The relationships to poverty and unemployment are stronger in central Canadian forest dependent CSDs than nationally as well. In the Prairie provinces, the relationship between forest reliance and unemployment and median family income is more negative than the national average. In Atlantic Canada and British Columbia, forest reliance appears to be associated with more positive outcomes for rural communities: forest reliance is associated with decreased poverty and increased median family income in each case. As well, there is no significant effect of forest reliance on unemployment in these regions.

What drives this regional variation? One explanation may lie in economic segmentation theory and regional variation in the sectors that comprise the forest industry. Overdevest and Green's (1995) discussion of core-periphery relations (Averitt 1968) is one example of this type of research in the United States. Disaggregating the forest industry into its four components-pulp, lumber, logging, and services--reveals implications for well-being that may be disguised by the general `forest dependence' label. Overdevest and Green (1995) identify the pulp sector as a core sector, with generally positive outcomes, with the others as periphery. In our data, forest services consists of consulting and contracting services, such as tree planting.

Table 3 illustrates the between-sector variation in the relationship between forest reliance and well being. Consistent with previous work examining core-periphery theories, pulp is advantageous relative to other forest sectors: it is the only sector not associated with lower educational attainment or higher unemployment (pulp dependent CSDs are essentially no different from average Canadian rural CSDs for these variables). As well, pulp employment bears a robust positive relationship to median family income, while the relationship for all of the other sectors is significantly negative. As such, pulp reliance presents itself in contrast to the other forestry sectors, which are all quite similar to each other: logging, services, and lumber are associated with lower educational attainment, higher poverty and unemployment, and lower income.

Table 3. Well-Being and Forest Reliance, by Sector

Are the above findings are general across Canada, or do they vary by region and the specific characteristics of places and characteristic products? The remainder of this paper focuses on median family income, educational attainment, and unemployment rates and explore whether the relationship between forest reliance (and its component sectors) and well being varies by region. Recall that the effects of overall forest dependence on income in particular vary between regions, ranging from negative in Central Canada (r= -.152) to quite positive in British Columbia (r= .178). Table four examines whether the forest industry differs regionally as to the performance of its component sectors. We have already seen that nationally, outcomes such as employment, income, and human capital differ according to the particular type of forest dependence observed. Do these relationships vary by region as well: does what constitutes a core-periphery relationship vary according to place, or is it primarily a matter of regional differences in the makeup of the industry (i.e., is there proportionately more pulp dependence in British Columbia, which, as a higher wage sector, accounts for a more positive overall effect), or are outcomes better for most sectors?

Table 4. Well-Being and Forest Reliance, by Forest Sector and Region

The top line in each row of Table 4 illustrates regional variation in forest sector representation. Nationally, the greatest share of forestry employment is in lumber (42%), followed by logging (30%). Pulp (16%) and forestry services (12%) represent much smaller shares of employment. These figures vary by region: the Atlantic provinces are less dependent on lumber (32%); the Central region is proportionally more dependent on lumber; the Prairie region and the North are reliant on forest services. British Columbia is less pulp dependent than the national average, reflecting the higher value of its round wood products.

How do each of these sectors relate to well-being, and are there regional differences in the relationship? We search for deviation between the regional and national level performance of forest sectors for community well being. It appears that there is a great deal of regional variation in the effect of sector reliance on well being. Logging reliance (proportionally higher in Atlantic Canada) results in especially low educational attainment in central Canada, with proportionally higher education attainment in the Prairie region and British Columbia. The relationship between logging reliance and family income is especially negative (more than the national average) in Atlantic and Central Canada, and is positive (higher employment in logging = higher income) in British Columbia. Logging strongly fosters unemployment in Central Canada, but there is no relationship in British Columbia. In short, logging has disproportionately negative outcomes in Central Canada, and positive outcomes in British Columbia.

In forest services, there is a strong negative effect on educational attainment in Central Canada. The negative effect on median family income is also strong in Central Canada, and in British Columbia. The relationship to unemployment is highest in Prairie and Central Canada, but quite weak in Atlantic Canada. In sum, forest services are associated with the most negative outcomes in Central Canada, and (relatively speaking) more positive outcomes in Atlantic Canada. As well, the positive effects of logging sectors on the well-being of rural British Columbia does not extend to forest services.

Lumber reliance appears to be another example of underperformance in central Canada: lumber reliance is associated with lower educational attainment, lower income, and higher unemployment than is true nationally. It appears to have slightly better-than-the-national-average outcomes in the Prairie region, producing higher educational attainment and median family income in particular. The relationship of lumber reliance in British Columbia closely parallels the national averages.

Nationally, pulp reliance has the most positive outcomes of any of the forest sectors. More than the other forest sectors, this appears to be a general, rather than regional effect (with the possible exception of British Columbia): there are no strong regional differences in the relationship between pulp reliance and educational attainment. Income in pulp reliant CSDs is proportionately higher in British Columbia, and unemployment is proportionately lower.

Summary and Conclusions

Despite the relative importance of the forest industry to rural life (and as an economic development strategy with widespread political support), little research has been conducted on the relationship between forest reliance and human well-being in Canada. Rather, the policy assumption has been that increased forest harvest and processing is a viable strategy for rural economic development. Our findings suggest that these questions deserve further scrutiny; in many parts of rural Canada, reliance on forest harvest and/or secondary processing is associated with negative economic and social outcomes compared with other rural CSDs. However, the overall relationship between forest reliance and human well-being is neither definitively positive, nor negative. The picture clarifies when national data are disaggregated by region: forestry is associated with decreased incomes in the Central region, increased income in British Columbia, and in other regions there is no significant positive or negative effect. Part of this variation is attributable to regional sector variation within the forest industry (i.e., regional differences in pulp reliance, consistently associated with higher incomes). However, it appears that this is not wholly the case, that there is regional variation in the performance of these sectors themselves.

References

Averitt, R.T. 1968. The Dual Economy: The Dynamics of American Industry Structure. New York: Norton.

Beckley, T.M. and T.M. Burkosky. 1999. Social indicator approaches to assessing and monitoring forest community sustainability. Information Report NOR-X-360. Edmonton, AB: Northern Forestry Centre, Canadian Forest Service.

Beckley, T. M. 1998. The nestedness of forest dependence: A conceptual framework and empirical exploration. Society and Natural Resources. 11(2):101-120.

Bliss, J, C. Bailer, G. Howze, and L. Teeter. 1992. "Timber Dependency in the American South." Presented at the 8^th World Congress for Rural Sociology, University Park, PA.

Canadian Council of Forest Ministers. 1998. National Forest Strategy: Sustainable Forests, A Canadian Commitment. Natural Resources Canada.

Cook, A. K. 1995. "Increasing Poverty in Timber-Dependent Areas in Western Washington." Society and Natural Resources 8:97-109.

Drielsma, J.H. 1984. "The Influence of Forest-Based Industries on Rural Communities." Dissertation, Yale University.

Fisher, D. 2001. "Resource Dependency and Rural Poverty: Rural Areas in the United States and Japan." Rural Sociology 66(2):181-202.

Force, J.H., GE Matchless, L. Zhang, and A. Kearney. 1993. "The Relationship Between Timber Production, Local Historical Events, and Community Social Change. A Qualitative Case Study." Forest Science 39(4):722-742.

Howze, G., C. Bailey, J. Bliss, and L. Teeter. 1993. "Regional Comparisons of Timber Dependency: The Northwest and the Southeast." Paper presented at the Annual Meetings of the Rural Sociological Society, August 7-10, Orlando, FL.

Humphrey, C., Berardi, G., Fortman, L., Geisler, C., Johnson, C., Kussel, J., Lee, R., Macinko, S., Schulman, M, and P. West. 1993. Theories in the Study of Natural Resource Dependent Communities and Persistent Rural Poverty in the United States. In Persistent Poverty in Rural America. Boulder, CO: Rural Studies Series, Westview.

Johnson, T.G., and Stallman, J. 1994. Human Capital Investment in Resource-Dominated Economies. Society and Natural Resources 7(3): 221-233.

Kaufman, H.F., and L.C. Kaufman. 1946. Toward the Stabilization and Enrichment of a Forest Community: The Montana Study. Missoula, MT: Univ. of Montana.

Nord, M. 1994. "Natural Resources and Persistent Rural Poverty: In Search of the Nexus." Society and Natural Resources 7:205-220.

Parkins, J., R. Stedman, and T. Beckley. 2002. "Forest Sector Dependence and Community Well Being in Canada: A Structural Equation Model for New Brunswick and British Columbia." Unpublished manuscript.

Parkins, J., and T. Beckley. 2001. Monitoring Community Sustainability in the Foothills Model Forest: A Social Indicators Approach. Information Report M-X-211E, Socioeconomic Research Unit, Canadian Forest Service. Fredericton, New Brunswick.

Parkins, J., Stedman, R.T., and J. Varghese. 2001. Moving towards locally defined indicators of sustainability in forest-based communities: A mixed-method approach. Social Indicators Research. 56(1): 43-72.

Power, T.M. 1996. Lost Landscapes and Failed Economics: The Search for a Value of Place. Washington, DC: Westview.

Overdevest, C. and G.P. Green. 1995. "Forest Dependent Community Well-Being: A Segmented Marked Approach." Society and Natural Resources 8:111-131.

Randall J. E. and R. G. Ironside. 1996. "Communities on the Edge: An Economic Geography of Resource-dependent Communities in Canada." The Canadian Geographer 40(1):17-35.

Statistics Canada. 1998. 1996 Census of Canada. Ottawa, ON: Statistics Canada

¹ The Pennsylvania State University, Dept. of Agri. Econ. and Rural Sociology, 111 Armsby Building, University Park, PA 16802, USA. [email protected]