Corps de l’article

In 1892, Alfred Marshall reflected on how, in business, the “tendency to variation is a chief cause of progress.” Explaining patterns of variation was, however, no easy task. As Marshall (1932: 205) remarked: “Every locality has incidents of its own which affect in various ways the methods of . . . every class of business that is carried on in it.” To complicate matters further, while in some trades, “the possible variations are confined within narrow limits,” in others “there can be great variations.” Although the processes of globalization have, by common consent, worked to undermine the stable employment relationships that characterized the post-1945 era, explaining patterns of variation remains a key task in industrial relations. Despite attempts to depict the current world order as “a borderless economy” (Ohmae 1995: 64), spatial barriers, national borders and local circumstance remain industrial and economic realities. For instance, one recent neo-liberal study of international trade (Fujita, Krugman and Venables 1999: 325) notes: “The real world is anything but seamless: It remains separated by oceans and deserts, by cultural and language differences, and by national boundaries that continue to impose substantial practical obstacles to trade.” This same study (1999: 239) also concluded that, despite trade liberalization, the United States-Canadian border “imposes barriers to arbitrage comparable to 1,700 miles of physical space.” Such observations reinforce the theoretical contributions made by radical economic geographers since the 1980s (Herod 1991, 1994, 1997, 1998, 2001; Massey 1994; Harvey 1982, 1989, 2001; Peck 1996; Smith 1984). Although these theorists are divided over the relative contributions of capital and labour to the construction of what Harvey (1982: 421) calls the “historical geography of capitalism,” they are broadly united in conceiving the latter as “a social process” whereby “productive forces and social relations . . . exist as particular spatial configurations.”

This study explores the ways in which “particular spatial configurations” have shaped the use of contactors in the export coalfields of Queensland (Australia) and western Canada since the late 1960s. During these years, Central Queensland and western Canadian producers assumed a dominant position in the international trade in the high grade coking coal used in steel making. By 2000–01, Central Queensland was exporting a record 76 million tonnes, their only serious competition the 27 million tonnes supplied by Canada (Coal Services 2001: 1; Coal Association of Canada 2003: 1). While the demand for coking coal from East Asia’s steel mills underpinned the expansion of the seaborne trade, the resultant production system rested on a complex set of social relationships, whereby vast sums of capital and thousands of workers and their families were settled in relatively remote locations. At all times in this system’s creation and maintenance, national and provincial governments played a decisive role. Not only did they expend billions of dollars on transport and social infrastructure, governments also provided the labour relations framework within which employment terms were negotiated. For miners and their unions, the issue of outsourcing was of particular significance, given the lack of alternative sources of employment. Not surprisingly, the various coal unions made the regulation of contractors an early objective in both regions, imposing almost identical restrictions on management.

If high coal prices and labour shortages initially provided a favourable circumstance for regulating contractors, a sustained fall in prices, which dropped from US$75 per tonne in 1983 to US$37.60 in 2000, undermined the industry’s earlier prosperity. As part of their response to falling prices, after 1996 Central Queensland’s coal owners repudiated previously negotiated agreements restricting their use of contractors. In the wake of these repudiations, the number of contractors rose from insignificant levels to more than one-third of the workforce (QDNRM 2003). By contrast, Canadian pits continued their historic reliance on employee miners. Given that both Queensland and Canadian mining operations were subject to the same global pressures, the key question is: to what can we ascribe variations in employers’ labour relations strategies with regard to use of contractors after 1996? In exploring answers to this question, this article finds the analytic framework provided by Harvey (1982, 1989, 1991) particularly useful. While others (Peet 1983; Herod 1991, 1994) have emphasized capital’s mobility and capacity for “flight,” Harvey has highlighted the ways in which productive capital, as well as labour, is geographically embedded. As Harvey (1989: 234) has observed: “When either capitalists or workers have important assets fixed and immobile in space, then neither side is in a good position to use powers of geographical mobility against the other.” Such insights have particular relevance to this study on two grounds. Firstly, in coal mining we are dealing with an industry where capital investments are physically fixed in the ground. Secondly, the crisis that beset the industry during the 1990s was the result not just of prior economic over-expansion, but also of geographical over-expansion, as mining was extended into what were, at best, marginal fields.

This article argues that an understanding of the strategies pursued by coal owners requires, first and foremost, a comprehension of the different historical geographies in which they found themselves. In Central Queensland, producers retained a position of relative locational advantage in the international coal trade. In these circumstances, in situ solutions were sought. The central feature of this Australian response was the successful utilization of the Workplace Relations Act, 1996 (WRA) to reshape local labour markets and industrial practices. By contrast, Canadian producers primarily reacted to the crisis by engaging in a process of geographic and organizational concentration that involved the abandonment of virtually all export operations outside British Columbia’s Elk Valley. Although, for miners outside the Elk Valley, this process involved a loss of employment, for those within the Valley it meant the retention of their strong local bargaining position.

An “Unobtainable Dream”: Contractors and the Historical Geography of Labour Relations

Addressing a United States symposium on contracting, Perry (1997: 532) remarked that: “The outsourcing of work either diffuses or diminishes union membership.” Others have likewise concluded that the ability of unions to counteract outsourcing and contracting is vital to the maintenance of effective collective bargaining patterns. In Canada, for example, Verge (2000: 281) has noted that corporate use of “networks of micro-enterprises” to reorganize production “is certainly bound to undermine the traditional collective bargaining regime.” At times, however, there has been a tendency to view the increase in outsourcing as an inevitable, universal trend. Two United States’ commentators have, for instance, noted: “There was a time, not so long ago, when the work of firms was done by employees of the firm. . . . For more and more working men and women, such a stable, long-term employment relationship with decent wages and benefits has become an unobtainable dream” (Hiatt and Jackson 1997: 488).

The growing literature of radical human and economic geography, highlighted in the introduction to this article, provides a useful antidote to such determinism. As Peck (1996: 266) remarked, “regimes of labour regulation have a tendency to take on distinctive local forms.” In taking this stance, these scholars are not arguing, as Massey (1994: 62) has emphasized, that each spatial distribution (nation, region, city) has its “own autonomous existence.” Instead, to cite Peck again (1996: 250), it is recognized that “local fortunes remain immensely vulnerable to macroeconomic forces.” It is also recognized that while the same general tendencies may operate differently in different places, they continue to be “both embedded in and part of wider regulatory systems” (Peck 1996: 109). The point, in short, is not to substitute a conceptual framework in which all outcomes are shaped by global factors with one in which locality has total primacy. Rather, it is an attempt to see geography itself as historically contingent. The “spatial,” as Massey (1994: 2) has observed, is not “some absolute independent dimension,” but instead “social relations stretched out.”

If radical economic geographers share a common belief that general historic forces tend to play out differently in different spatial contexts, there is little unity of thought when it comes to identifying the general forces at play. There is disagreement as to whether labour campaigns conducted at the local level are progressive or regressive strategies. In the United States, Herod (1991, 1994), Peet (1983) and Johns (1998) have all argued that “capital flight” to non-union, low-wage areas has been the key factor in the deindustrialization of the old labour “heartlands.” To rectify this situation, Herod (2001: 409) believes, organized labour needs to adopt “a model which focuses not upon the global scale of worker organization but, instead, upon the very local scale of organization.” Both Herod’s “deindustrialization” thesis and his support for a local model of organizing have, however, been subject to refutation. Martin, Sunley and Wills (1994: 62) have concluded that the empirical evidence does not support the thesis that “capital flight” is a “major cause” of either “regional or local deindustrialization.” They are also dismissive of Herod’s call for “place-specific campaigns to defend local jobs,” observing that such campaigns will merely result in further job competition “between different groups of workers” (Martin, Sunley and Wills 1994: 64). Echoing such sentiments, Peck (1996: 255) has remarked: “Localization is critically connected to the weakening of labour’s bargaining strength.”

Where Herod has focused on the potential for workers to remould their own geography, Harvey (2001: 323) has emphasized how capitalism strives “to create a social and physical landscape . . . requisite to its own needs at a particular point in time.” This, Harvey notes, can only occur if capital is taken out of circulation and fixed in place. Once embodied in place, however, capital becomes exposed to devaluation. Whatever the cause, “devaluation is always specific to a particular place and time,” with general crises resulting from aggregations of local devaluations (Harvey 1982: 425). For Harvey, the desire to escape devaluation provides the key to understanding the behaviour of both capital and labour. In each crisis, individual elements of capital and labour are confronted with the choice—either “stay in place and fight” or “run for the safest havens” (Harvey 1982: 441). Such “fight or flight” decisions are shaped, in large part, by the actions of the state. While the state may “bring forward” the process of capital devaluation through tough monetary or fiscal policies, it not infrequently opts to transfer the costs of devaluation from capital to labour by implementing labour laws favouring the former (Harvey 1982: 449).

While, as previously indicated, Harvey’s analysis has particular relevance to the mining industry, his work has been criticized for its failure to properly consider the ways in which the spatially embodied manifestations of capital can be contested by labour. Massey (1994: 140), for example, has noted that in Harvey’s work “it seems, only capital can win.” In this regard, the recent work of Johns (1998) provides a useful addition to the literature. While sharing Harvey’s views that “workers cannot win” anything but short-term relief from locality struggles, Johns (1998: 268) postulates a model that prioritizes international solidarity campaigns whose principal aim is the reshaping of capital movements and investment decisions.

Investment Strategies and Geographic (Over)-Expansion

Coal is a low-value commodity that has historically been consumed close to the point of production. While 88 per cent of the world’s coal is still consumed locally (IEA 1998: I.221), from the 1960s both Central Queensland and western Canada were drawn into a global revolution in the consumption and production of coal. On the demand side, the key factor in this revolution was the appetite of the Japanese steel mills (JSM) and power utilities who, by 2001, were importing a record 155 million tonnes of product (Tex Report 2002). On the supply side, new market needs demanded new production systems, as underground mining proved incapable of producing large volumes at low cost. The key development on this front was the adoption of highly capitalized opencut mining, techniques perfected by the US-based Utah Corporation. In 1965, this latter company signed long-term contracts to provide coking coal to the JSM from Central Queensland’s Bowen Basin. In western Canada, the lure of long-term contracts was even more significant in enticing producers into the market, given that mining areas were over 1,100 kilometres from deep-water. After Kaiser Resources’ negotiation, in 1968, of the first long-term contract to supply the JSM from the Elk Valley of British Columbia (BC), western Canada’s coal industry experienced a rapid revival. Whereas, in 1968, BC’s output was just 900,000 tonnes, by 1998 the western provinces of BC and Alberta were shipping 34.2 million tonnes (Coal Association of Canada 1998).

The creation of new patterns of production and consumption in the Pacific Basin coal rested, in the first instance, upon the implementation of investment strategies forged between buyers, producers and provincial and federal governments. Significantly, organized labour had no role in this process. In Central Queensland, successive conservative governments engaged in what Galligan (1989: 70) has described as “development boosterism,” underwriting the construction of new railways, ports and townships. In western Canada, the provision of state-funded infrastructure was vital, given the region’s locational disadvantage. To subsidize western Canada’s entry into the market in the face of spatial barriers, the Canadian and BC governments expended in excess of Can$2 billion to create “one of the finest transportation and port systems in the world” (Marshall 1991: 43). In undertaking such expenditure, in 1976 the BC Minister for Economic Development advised Japanese buyers that federal and provincial governments were determined “to keep the price of Canadian coal exports competitive” (cited, Dames and Moore 1977: 37). Much of this state funding was used to foster mining in the isolated Tumbler Ridge region of North-East BC, where the Quinette and Bullmoose mines began production in the early 1980s. However, the remoteness of these mines, and of subsequent operations developed in the Grande Cache region of the Alberta foothills, meant that their survival—even with state support—was dependent upon the JSM’s willingness to pay a premium for their output.

Although developmental costs in both western Canada and Central Queensland were substantially underwritten by the state, there remained a competitive pecking order based upon locational advantage. In has been estimated (IEA 1998: I.166; McCloseky 1997: 59) that in 1997 the combined rail, port and shipping cost of transporting a tonne of coal from Central Queensland, via Dalrymple Bay, to Japan amounted to US$22.22. By contrast, on average it cost western Canadian producers, shipping through Vancouver’s Roberts Bank facility, US$28.34 to transport a tonne of coal to the same market. For producers in North East BC and Alberta exporting through the northern port of Prince Rupert the costs would have been even higher. To retain western Canadian producers in the market, Japanese buyers deliberately placed them “at the high cost end of the Japanese supply portfolio” (Marshall 1991: 37). As a result, Canadian operators received on average US$10–15 per tonne more than their Australian counterparts. These pricing practices created a dual dichotomy within the spatial structures of the Pacific Rim coal trade. Not only were Canadian producers subsidized in order to offset locational disadvantage, there also existed within Canada what Marshall (1991: 37) has referred to as a “two price system,” as producers around Tumbler Ridge and in the Alberta foothills obtained higher prices than those in the Elk Valley.

The combined effect of state infrastructure support and differential pricing had a number of long-term consequences for labour relations in both countries. Firstly, it created an initial period of high returns; a fact demonstrated when, in 1977, Utah announced the largest profit ever recorded in Australian financial history (Galligan 1989: 15). In such circumstances, there was little incentive to engage in industrial confrontations with mine-site workforces that would interrupt supply. Secondly, this process resulted, as the Japanese buyers intended, in an increasing over-supply of the market as a result of economic and geographic over-expansion. Finally, the process of geographic expansion ensured that, if price differentials were discontinued, then those operations based in the Alberta foothills and around Tumbler Ridge in British Columbia would be particularly vulnerable.

“The Most Significant Constraint”: Permanent Workforces and Labour Organization

While organized labour was excluded from the investment processes that provided the economic framework within which the export coal industry developed, the industrial and social problems associated with creating new workforces soon overshadowed all other operational issues. As early as 1976, one Canadian study observed that “manpower recruitment and training” were “the most significant constraint to the development of the coal resource” (Technical Committee 1976: 15). In both Central Queensland and western Canadian the obstacles to overcoming these constraints were considerable. In 1966, the population of the Central Queensland coalfields area was less than 8,000, while the town of Blackwater, adjacent to Utah’s first mine, was described “as a whistle-stop of the lowest order” (cited Galligan 1989: 83). In western Canada, the problems associated with building and maintaining a skilled labour force was, if anything, even greater. In North East BC, mining development occurred in a virtual wilderness. Even in the historic Elk Valley mining centre of Michel there were, in 1961, a mere 417 residents (Hedum Menzies 1969: 67–68).

In dealing with the constraints imposed by labour shortages, Canadian and Queensland producers pursued an identical strategy, working with local governments to develop the infrastructure capable of supporting large permanent workforces. As one Canadian study observed, “every mine asked the communities to develop more land and infrastructure for housing to meet increases in the number of employees required to produce the additional coal. Each company wanted quality, affordable housing available to attract a good workforce” (Marshall 1991: 74).

Although, in both regions, the costly process of assembling resident workforces can be attributed, in part, to remoteness, the latter cannot fully explain the owners’ actions. For, after all, mining, whether for coal or hard-rock, has long been associated with various forms of contract labour and impermanent shanty towns. That this did not occur in the developmental stage of the export industry can be attributed to three factors, in addition to labour market remoteness. Principally, in underwriting development, both Canadian and Australian governments had done so with the explicit intention, as one Canadian study noted, of fostering “significant regional development based on coal exports” (Technical Committee 1976: 349). For these governments, the establishment of permanent townships was the quid pro quo for their investment in the industry. As well, opencut mining for the supply of export markets was entering virgin territory, both technically and geographically. Given the costs and expertise involved, there were few companies, other than the actual producers, who possessed the capacity to assemble mining equipment and associated labour on a large scale in the required locations. Additionally, owners soon found themselves dealing with entrenched labour movements who soon demonstrated an ability to shape industrial developments.

Except for the wilderness areas of North East British Columbia, the export sector in both Queensland and Canada was superimposed over a landscape where underground miners, and their unions, had long been active. In Central Queensland, the miners who laboured in the underground pits that dotted the Bowen Basin’s fringes were “renowned” for their “staunch union loyalty” (Thomas 1986: 99). Similarly, coal mining in the Alberta-British Columbia border regions produced what Langford and Frazer (2002: 49) have described as a “strong, indigenous socialist workers movement.” In Queensland, these labour traditions were primarily embodied in the Miners’ Federation, now the Construction Forestry Mining and Energy Union (CFMEU). Prior to 1996, the latter’s organizing task benefited from the fact that all Australian employment operated within the confines of federal or state “awards.” Between 1946 and 1995 coal miners found themselves in a particularly privileged position, as a separate Coal Industry Tribunal oversaw their conditions. This strength, however, was also the Achilles heel of Australian unions, given that their right to represent workers was embodied within this framework. Meanwhile, in western Canada, the United Mineworkers of America (UMWA), the United Steelworkers of America (USWA) and the International Union of Operating Engineers (IUOE) all sought the loyalty of coal miners. Of these latter unions, the USWA proved the most successful. Despite falling employment, as late as 2001 the USWA still represented two-thirds of the sector’s unionized workforce (USWA 2001). Collectively, the Canadian coal unions imposed standardized employment conditions across the export pits. While the Elk Valley’s Line Creek mine began in 1981 on a “staff employment” basis, it was soon unionized. Similarly, although Fording’s Greenhills mine reopened in 1993 as a non-union concern, employment was regulated by an Employee Handbook that mirrored arrangements elsewhere (Interviews 2002). For all three unions there was, as one USWA representative emphasized, “total union opposition to the contracting out of ‘bread and butter’ areas of employment” (Hunt 2001).

Securing Regulation: Contractors and Collective Agreement Provisions to 1996

If the “most significant constraint” confronting both Central Queensland and western Canada producers during the initial export era was a shortage of labour, this problem was particularly pronounced during unscheduled breakdowns, “spike” maintenance periods and infrastructure construction. In such circumstances, employers quickly realized the benefits of negotiating industrial agreements that would guarantee them access to contractors during such periods of peak demand. In Central Queensland, the first such agreement occurred in July 1974, when Utah agreed that any use of contractors would be preceded by discussions with union representatives. This was followed by the negotiation of a framework agreement with the Queensland Coal Owners’ Association—the Queensland Maintenance and Minor Construction Agreement (QBOR 1975). Under the agreement’s terms, there were only three circumstances where contractors could be used. Firstly, there was work requiring “a significant short-term increase in manning.” Then, there was work that involved the use of technologies “not readily available.” Thirdly, there were conditions where “particular skills” were required on “a limited basis.” In all circumstances, union representatives were to be given six weeks notice. In addition, contractors’ employees were obliged to join “the same union as their employed counterparts” (BHP-Utah 1988). If the Agreement ensured that any contracting out of work occurred within a regulated environment, it was also effective in keeping the use of contractors to a minimum. As late as 1996, the number of contractors and their employees working within the region’s opencut mines amounted to a mere 416 individuals, compared to an employee workforce of 6,665 (QDNRM 2003).

In western Canada, coal unions were also successful in imposing, through collective bargaining, a regulatory regime that closely resembled the one established in Central Queensland. While pre-1990 copies of collective agreements have, unfortunately, not been preserved, oral testimony suggests that more recent agreements replicate earlier understandings. Between 1994 and 2002, for example, the collective agreements negotiated at the Fording River (1996, 2001), Quinette (1994, 1997), Line Creek (1997, 2001), and Elkview mines (1994, 1997, 2000) in BC, and the Smokey River (1995) and Luscar (Cardinal River 1998) mines in Alberta, all contained provisions identical, or similar to, the following: “The Company and the Union agree that contracting out of work performed by employee in the bargaining unit will not be done for the purpose of laying off, demoting, terminating, or deferring the recall of bargaining unit employees” (Quinette 1995: 5).

In addition to such motherhood statements, all agreements provided for regular reviews (normally monthly) of all contracting out proposals by a “contracting out committee,” or equivalent. These committees were entrusted with the task of considering, in the words of successive Fording River agreements (1996, 2001), “the use of contractors at the mine site with a view to promoting their mutual interest of placing primary reliance on the Company’s employees consistent with sound business practice.” Where it was determined that specific tasks would be contracted out, unions were to receive appropriate advance notice (normally 30 days), including details about the proposed contractor and the anticipated number of the contractor’s employees who would be on-site (Luscar 2000; Smokey River 1995). At some mines, such as Alberta’s Smokey River (1995: 1.01), this process of notification included a further discussion of “all options” for “directing the work towards employees.”

Divergence in Patterns of Regulation, 1996–2002: Amortization and Geographic Concentration in western Canada

Although, prior to 1996, there were marked similarities in the negotiated agreements that determined how contractors were employed in both Central Queensland and western Canada, from this point practices rapidly diverged, as Australian producers sought “unfettered use of contractors” while their Canadian counterparts did not. In large part, this divergence reflected the different geographic placement of each region in the Pacific Rim coal trade, and a corresponding variance in coal owner strategies. As coal prices fell by 18 per cent during 1996, all producers were forced to review how this placement affected their operations. For western Canadian producers, the fundamental problem remained distance from market. As one industry journal reported in 1998: “Western Canadian mines have demonstrated high productivity, but are at a competitive disadvantage to Australian mines because of the long distances between mines and ports” (Carrington 1998: 66). In such circumstances, the industry’s woes could not be remedied simply through productivity improvements or labour reforms. Indeed, Canada’s placement at the high cost end of supply occurred despite the fact that, by 1997, mine-site labour costs averaged just US$4.60 compared to an Australian average of US$11.90 in 1996 (IEA 1998: I 204–05). There were, however, limited possibilities for further gains in this area. On the productivity front, BC producers lagged behind their Queensland rivals, despite the annual output of each BC miner increasing from 4,310 to 6,619 tonnes between 1990 and 1996 (BCMEM 2003). Again, geography worked against Canadian operators, as Queensland’s superior productivity (where each miners’ annual output averaged 9,221 tonnes in 1995–96) largely reflected the use of large-capacity draglines. While such machinery could be used with ease on Queensland’s central plains, on the mountaintops of western Canada they could be rarely utilized, forcing owners to resort to less efficient truck-shovel methods. On the wages front, Canadian producers also had little room to manoeuvre, given that, in 1996, the US$40,300 received by the average Canadian coal miner was well below the US$58,400 paid to his Australian counterpart (IEA 1998: I 204–05). Here, Canadian producers were caught in a hard place, as the need to pay a “premium” to maintain suitable workers in remote locations limited their capacity to reduce labour costs in order to offset locational disadvantage.

As the export coal sector increasingly became a loss-making venture in western Canada, both producers and provincial governments faced the total amortization of their capital investment in the industry. This was particularly the case in the Tumbler Ridge and Grande Cache regions, as the Bullmoose, Quinette and Gregg River mines saw the expiry of long-term contracts containing payments “significantly in excess of market prices” (Luscar 1999: 13). The inevitable result was the abandonment of those operations with the highest transport costs, and an organizational and geographic concentration of the surviving producers in the Elk Valley. Between 1999 and 2002 all export operations outside the Elk Valley closed, resulting in the loss of over 1,500 jobs (USWA 2001). In order to salvage their remaining coal assets, in February 2003 the three surviving Canadian operators—Fording Coal, Tech Cominco and Luscar—agreed to the division of the Canadian industry into two cartels. Under the terms of a “Non-competition agreement” approved by Alberta’s Court of the Queen’s Bench, the newly-established Fording Coal Partnership acquired all of western Canada’s coking coal assets, with the right to manage these resources free of competition from its former rivals. In consequence, the Partnership secured the Elk Valley’s entire annual output of 25 million tonnes (Fording Coal 2003).

The concentration of the western Canadian industry on the Elk Valley had a major regional economic impact. In commenting on the closure of Alberta’s Smokey River mine, for example, the mayor of nearby Grande Cache observed: “The mine is Grande Cache’s biggest employer by far, so this is going to be devastating” (Edmonton Journal 27 February 2000: A7). However, for those miners who retained their jobs in the Elk Valley, the process of geographic concentration actually strengthened their labour market placement. By mid-2002, employment in the Valley’s five mines stood at over 2,600—not far short of that for the total BC coal industry two years before (Interviews 2002; PricewaterhouseCoopers 2001: 10). Given that the Valley contained just 13–14,000 people, this presented significant recruitment problems. To compound matters, producers also faced competition for skilled maintenance workers in particular from both the hard-rock sector and Alberta’s lucrative “oil-sands” projects (Interviews 2002).

As western Canadian coal owners engaged in the difficult process of organizational and geographical concentration there was little interest in engaging in industrial conflicts over what were, for them, peripheral issues such as outsourcing, particularly given the tightness of the labour market in their remaining operational area. On the contrary, during the crisis period 1996–2001 all mining companies reiterated their support for historic labour practices. Thus a senior Executive of one company observed during 2001, that co-operative union-management relationships remained a “given” (Interview 2001a). Emphasizing this point, by 2001 all of the Elk Valley’s major unionized operations had renegotiated long-term collective agreements which replicated earlier restrictions on the use of contractors (Line Creek 2001; Fording River 2001; Elkview 2001). In addition, four of the five Elk Valley mines, including the non-union Greenhills operation, maintained a “no lay-off” policy during the worst of the downturn. Rather than resorting, as their Australian counterparts did, to widespread use of outsourcing, mine management negotiated job-sharing and leave arrangements with their workforces (Interviews 2002).

Divergence in Central Queensland: In Situ Employer Strategies and the Engagement of Contractors

Where western Canadian coal owners’ primary response to falling commodity prices involved the spatial reorganization of the industry in order to offset locational disadvantage, the problems and responses of Australian producers were radically different. As previously noted, in 1997 Central Queensland producers possessed, on average, a US$6.12 per tonne transport advantage over their Canadian competitors. However, by 1996–97, average mine-site labour costs in Australia were US$7.20 per tonne higher than they were in Canada. This effectively meant that the entire locational advantage of Australian producers had been “captured” by the workforce. While, for employers, this could be overlooked during periods of high price, their forbearance disappeared as prices fell. Given Central Queensland’s continuing locational advantage, there was little consideration given to large-scale mine closures in the opencut sector. These remained extremely valuable assets that needed to be defended through in situ strategies that focused on a lowering of labour costs.

As producers contemplated a major overhaul of their work practices, an increase usage of contractors offered three perceived benefits. It reduced direct capital costs, since, as the head of Peabody Resources noted, the practice normally involved “contractors supplying their own equipment” (Humphries 1998: 3). In contrast to the industry’s developmental era, there were now a number of companies willing to undertake large-scale contract mining work. As one such firm, Roche Mining, noted in its corporate magazine during mid-2000: “Roche has actively targeted the coal sector, which represents a big market with a small contractor penetration” (Newsfront June 2000: 5). Contracting out also provided a mechanism for the introduction of new work arrangements such as longer shifts and reduced crew levels. Perhaps most significantly it provided a “change agent” for forcing industrial change on the wider workforce. Nevertheless, any extension in the use of contractors in Central Queensland remained problematic while those factors that underpinned the Maintenance and Minor Construction Agreement—entrenched union presence, centralized industrial relations legislation that guaranteed union coverage and the tightness of the regional labour market—remained in force.

Given the capacity of mining unions to block extensions in the use of contractors within the industrial framework that existed prior to 1996, this issue became inextricably linked to a campaign by Queensland’s coal owners for dramatic changes to national labour laws. Among the coal owners, CRA/Rio Tinto played a particularly significant role. In 1991, for example, CRA’s Leigh Clifford advised other coal company executives that “a major change was needed in Queensland industrial policy.” Such changes, he argued, were required to shift the focus “away from centralized negotiation of industrial issues at state and national level” to “individual site discussions” (Grubb 1991). In taking this stance, Clifford was reflecting his company’s preference for workplaces where trade unions were excluded from any meaningful role. As Peetz (2002: 252) has noted, in Australia the implementation of a “decollectivist model” emerged “initially through the operations of companies in the CRA/Rio Tinto group.” During the early 1990s, this model was forcefully implemented in a number of the company’s non-coal operations, most notably New Zealand’s Tiwai Point refinery and Australia’s Bell Bay and Boyne Island aluminium smelters.

By April 1992, CRA’s insistence on the need for radical change had won the support of the Queensland Coal Owners’ Industrial Organization, who accepted a five-year strategy to bring about an “uncentralized and site focused industrial relations model within the industry” (IRSAG 1994: 1). Significantly, an increase in “contractors and outsourcing” was identified as one of the sought after “changes in behaviour and resultant working arrangements” (IRSAG 1994: 16). To implement this strategy, it was agreed that the industry’s peak councils should actively support radical legislative change. In 1993, the coal owners’ federal body, the Australian Coal Association, duly took up the baton, informing the federal government that: “Reform is urgently needed in the coal industry” (ACA 1993). Intellectual endorsement for radical change also came from the conservative H. R. Nichols Society. In one of its earliest publications, the Society noted that producers’ capacity to respond to the “squeeze action between rising costs and falling prices” was constrained by “peace at any price” labour laws (Larratt 1988: 62). Only by removing these constraints, it was argued, could the coal industry prosper.

For Queensland’s coal owners, the passage of the WRA in December 1996 represented the vital breakthrough. As CRA’s historian (Ludeke 1996: 45) observed at the time: “The environment which will be introduced by the new law will challenge the foundations of union structures in Australia.” Such proved to be the case. On 10 January 1997, with the new act in its infancy, management at Rio Tinto’s Blair Athol mine declared that it would no longer require contractors’ employees to join a union (Bates 1997). In July 1997, management went further, giving “notice of our intention to withdraw from the Maintenance and Minor Construction Agreement” (McCrea 1997). Despite a series of bitter conflicts, by early 2001 the CFMEU was forced to concede defeat, agreeing to a new Certified Agreement which stipulated management’s right to use contractors “on any work as required” (AIRC 2001a).

Following Rio Tinto’s success, other employers were quick to follow suit. In January 2001, BHP Coal advised unions of its intention to enforce its “right to free and unfettered use of contractors” at all its Queensland operations (BHP Coal 2001). When the subsequent Enterprise Agreement was ratified, these words were replicated without modification (AIRC 2001b). No longer constrained by formal agreements, employers dramatically restructured their workforce through the engagement of contractors. As Table 1 indicates, between 1996 and 2002 the number of contractors and their employees engaged in Central Queensland’s opencut mines increased six-fold to 2,480. During the same period, the number of direct mine employees declined from 6,665 to 3,954.

Table 1

Employees and Contractors (and their Employees), Central Queensland’s Opencut Mines, 1996–2002

Employees and Contractors (and their Employees), Central Queensland’s Opencut Mines, 1996–2002
Source: Statistical Office, Queensland Department of Natural Resources and Mines.

-> Voir la liste des tableaux

Socially, the increased use of contractors appears to have contributed significantly to a sharp fall in the coalfields’ population, which fell from 42,571 to 40,789 between 1996 and 2001. By the 2001 census night, more than 28 percent of homes in the mining towns of Dysart and Middlemount were abandoned or otherwise unoccupied. In Collinsville and Blackwater, the percentages were 23.9 and 18.8 percent respectively (QDLGP 2001: 4, 21). Industrially, there is also little doubt that the removal of earlier restrictions on contractors assisted coal owners in reducing costs. Between 1995–96 and 1999–2000 total employment in the Queensland coal industry, including “on site” contractors, fell from 10,962 to 7,972, while saleable output per employee per year almost doubled, rising from 7,709 to 14,483 tonnes (Coal Services 2001: Table 1). While statistical evidence on mine-site labour costs remains unavailable at an industry level, Rio Tinto reported gains of US$6–7 per tonne at its NSW Hunter Valley No. 1 mine (Australian Mining 1999: 9). If such gains were, in fact, replicated throughout the industry, they would represent an employer recapture of the locational advantage surrendered to their workforce during the period of high price.

If the WRA enabled Central Queensland’s coal owners to create a new employment regime, employers were less successful in overcoming the union loyalties of the region’s workforce. As other barriers to the use of contractors diminished, these loyalties provided employers with their most significant obstacle to the unregulated use of outsourcing. This, in turn, appears to reflect the fact that contracting out in Central Queensland did not lead to the general displacement of the mining workforce, as miners opted to work for contractors rather than leave the industry. Thus, while official statistics do indicate a movement of displaced miners to coastal towns such as Mackay (QDLGP 2001), anecdotal evidence also suggests that a significant proportion, if not a majority, of those engaged by contractors were redundant coal miners (Interviews 2003). While accurate figures on union density rates among contractors’ employee are also unavailable, CFMEU activists employed by contractors have estimated that 85 per cent of such workers are union members (Interviews 2001b). Significantly, by 2000 it appeared as if the process of redundancy and contracting out initiated by the WRA was losing momentum. While the number of contractors employed in Central Queensland’s opencut mines more than doubled between 2000 and 2002 to reach a record 2,480, the number of directly employed miners also rose from 3,373 in 2000 to 4,203 a year later. Overall, the total engaged Queensland workforce rebounded strongly after 2000, growing by approximately 25 percent from 8,105 to 10,205. Partly reflecting this reversal, in the year to September 2002, productivity in Queensland coalmines fell by a record 13.78 per cent (QDNRM 2002). While it is unclear whether this marked productivity decline is associated with growing difficulties in managing contract miners, any early return to the pre-1996 reliance on directly engaged workforces appears unlikely.

Labour Strategies and Responses

While the different strategies pursued by Canadian and Queensland coal owners in reaction to falling prices reflected, in large part, their spatial placement in the Pacific Rim coal trade, organized labour’s responses remained embedded within national regulatory systems. In western Canada, the USWA’s principal concern during the post-1996 downturn was that there should be no surrender of the “cornerstones” of [business] unionism – job seniority, work demarcation lines, and restrictions on contracting (Hunt 2001). While USWA membership in the sector fell from 2,445 in 1996 to 1,210 in 2001 (USWA 2001), the union was prepared to acquiesce in mine closures rather than see Locals engage in concession bargaining (Hunt 2001). Thus, when Alberta’s Smokey River mine closed, the principal grievance of the USWA Local was the lack of “notification” that would have allowed miners to “look for a job” (Edmonton Journal 27 February 2000: A7). In part, this stance reflected the USWA’s view that issues of investment and economic policy were ones best handled through its affiliation to the New Democratic Party. This is not to say that the union was blind to the changes being wrought by global economic activities. Indeed, the union imposed a levy on all members for a Humanity Fund aimed at assisting workers in developing countries avoid a “race to the bottom” (Hunt 2001). In evaluating possible union responses to international “spatial competition,” however, Johns (1998: 258–9) would categorize this latter strategy as representing a “classic protectionist view,” given that its principal object is to protect local jobs by raising costs elsewhere. Such a viewpoint, she believes, remains fundamentally incapable of “challenging the spatial organization of global production (Johns 1998: 269).

Whereas the principal reaction of the USWA to their industry’s spatial reorganization was the protection of traditional collective bargaining standards and practices, the major response of Australian coal unions to the increased usage of contractors was also found in the industrial arena. In mid-1995 the CFMEU finally accepted that it was no longer in a position to exclude contactors from core mining activities, resolving instead “that an active recruitment campaign of contractors be immediately initiated (CFMEU 1995: 375). These campaigns resulted in the larger contracting companies being covered by industrial agreements, even under the adverse conditions applying under the WRA (see, for example, AIRC 2002). However, the industrial impact of these new agreements was diluted by their failure to contain any provisions dealing with the actual contracting out process. Thus, in negotiating an agreement with a contractor at the large Goonyella-Riverside mine, the union conceded him the unrestricted right “to contract out the functions such as major plant repairs, refurbishment and servicing, and component change-outs, drill and blast, and rehabilitation work” (AIRC 2002). While such agreements were, by themselves, incapable of restoring security of employment, the unions were even less successful in organizing solidarity campaigns at either the local or global level. Among miners forced into working for contractors, the lack of support from local politicians and shopkeepers was particularly galling (Interviews 2001). Similarly, although the CFMEU played a formative role in the Coalition of Rio Tinto Shareholders, and established alliances with international unions covering miners employed by Rio Tinto, these actions did little more than publicize Rio Tinto’s unenviable labour relations record (Bowden 2000: 378–9).

Perhaps the most surprising aspect of the responses of Central Queensland and western Canadian unions to the crises that beset their industry is their failure to engage with each other. Despite the fact that Queensland and Canada have been competitors in the export coal trade since the 1970s, there has been no significant trans-Pacific union dialogue. In consequence, unions in each region have remained ignorant of even the most basic industrial developments in the other. There has, as a result, been no discussion of joint industrial action to exploit employer vulnerability during annual price negotiations with buyers. Nor has there been any discussion of the ways in which Australian and Canadian unions can collaborate to influence investment and pricing arrangements.


In 1848, Marx and Engels (1951: 36) made a famous observation that capitalism dissolved “fixed, fast-frozen relations” to such an extent that: “All that is solid melts into air.” Yet, while their analysis foreshadowed much of the recent literature on capital mobility (Herod 1991, 1994, 2001), in exploring variations in the regulation of contractors in the export coal sectors of western Canada and Central Queensland this article supports Harvey’s contrary thesis (1982: 425, 441) that capital, once fixed in place, cannot be moved without exposure to devaluation. This general rule, it is argued, is of particular importance in mining. For, as Rio Tinto’s Chairman observed in 1999: “Once capacity has been created, it is very difficult to remove” (Wilson 1999: 3). In the current study, the difficulty in removing fixed capital, even during deteriorating economic circumstances, manifested itself in a process of geographic over-expansion. This geographic over-expansion did not stem, however, from the “invisible hand” of market forces but rather from deliberate investment decisions made by producers, governments and buyers from the 1960s onwards. Significantly, organized labour was not involved in this investment process. While the demand for coal from Japan’s steel mills provided the impetus for development, provincial governments in both western Canada and Queensland played the decisive role in overcoming the spatial barriers between coal resources and markets through their construction of ports, railways and social infrastructure. By subsidizing private development, however, state policies fostered the extension of mining into economically marginal fields. Although, in the short term, producer profitability was guaranteed, this geographic over-extension was at the root cause of the crisis of over-supply that was clearly evident by the 1990s.

The responses of the Queensland and Canadian export producers to the crises of over-supply differed profoundly according to differences in their special placement. For Canadian owners, the key issue was locational disadvantage, due to the distance of mining operations from deep-water ports. In consequence, Canadian owners reacted to falling prices by concentrating mining operations in the most favourable location, the Elk Valley. Given the scale of this restructuring exercise, employers had little incentive to engage in industrial battles over what were, for them, peripheral labour issues such as the use of contractors. As a result, previously negotiated restrictions on the use of contractors remained in force. By contrast, Queensland producers were forced to deal with the industrial consequences of superior location. In essence, Central Queensland employers had allowed, during the period of high profitability, their entire transport advantage vis-à-vis western Canada to accrue to their workforce via higher mine-site labour costs. To alter this situation required a radical transformation of employment relations. As a result, from 1992 Queensland coal owners became active supporters of a major overhaul of Australian labour laws; a campaign that was crowned with success following the passage of the WRA in December 1996. The decisive impact of the WRA in transforming labour relations is particularly evident in the use of contractors, as the numbers engaged rose from minuscule levels in 1996 to more one-third of the workforce just four years later.

The significance of spatial configurations in shaping employers’ labour strategies in both western Canada and Central Queensland emphasizes the need to conceptualize place not just as a local, physical manifestation but also as a set of social relations that play out simultaneously at all spatial levels, including the global and national. Although global processes play out differently in different national and local contexts, as Peck (1996: 250) notes, “many of the determinants of local success or failure continue to be exogenous not endogenous.” Thus, while organized labour was, in the current study, capable of exerting considerable influence over the ways in which employment relations were configured at the local level prior to 1996, these local strategies remained at the mercy of global forces. In these circumstances, while scholars such as Herod are correct in emphasizing the importance of local campaigns, there remains a need to transcend such mobilizations by linking them to wider struggles capable of influencing the investment processes through which capital takes material form. In this regard, Harvey’s analysis, which emphasizes the ways in which capital must remained fixed in place if it is not to be devalued, highlights the opportunities and challenges inherent in any place-based production system. For, as Harvey (2001: 360) observes: “The accumulation of capital has always been a profoundly geographical affair.” As the geography of capital becomes increasingly global, so too must be the geography of labour.