SECTOR Home | What's New | About SECTOR | Meetings | Publications | Discussion Forum | Contact Us | Links | Site Map

WP.162

Contents

• Foreword

• 1. Background and context

• 2. Results of the survey of the Australian mining industry (February 2000)

• 3. Implications of working time trends for the Australian mining industry

• 4. Conclusions
• References

Figures
1.1	Change in hours for full-time employees in Australia 1978-97
1.2	Change in the distribution of working hours in Australia, 1983-98: all employees
1.3	Change in the distribution of working hours in Australia, 1983-98: Males
1.4	Change in the distribution of working hours in Australia, 1983-98: Females
1.5	Changes in average hours of work: full time employees
1.6	Average weekly hours in mining in Australia and selected countries
1.7	Common provisions in collective and individual agreements
1.8	Hours provisions in registered collective and individual agreements
2.1	Days of operation: seven days operations and others
2.2	Operating hours of mines: 24 hour operations and others
2.3	Community setting of mines
2.4	Standard shift lengths in production
2.5	Standard shift lengths in maintenance
2.6	Shift lengths in production (excluding LDC sites)
2.7	Maximum number of consecutive shifts 12-hour sites (production)
2.8	Maximum number of consecutive shifts worked (maintenance)
2.9	Maximum number of consecutive shifts 12-hour sites (production)
2.10	Maximum number of compressed shifts (production only) excludingLDC sites
2.11	Average ordinary weekly hours (production)
2.12	Average ordinary weekly hours (maintenance)
2.13	Hours of work in 12-hour shift sites (production)
2.14	Overtime worked (production)
2.15	"Has the impact of the roster been assessed?" (Production and maintenance).

Tables
1.1	Hours of work in the mining industry, 1991 and 1996
2.1	Distribution of responding mines by state (in order of number of mines)
2.2	Major products produced by mines (all respondents)
2.3	Mining methods by state (per cent)
2.4	Times of operation by community setting (number of mines)
2.5	Standard shift lengths for production and maintenance employees
2.6	Shift length by mining method: production only
2.7	Type of roster cycle for production and maintenance employees
2.8	Ordinary hours by community setting: Production and maintenance
2.9	Overtime arrangements for production and maintenance employees
2.10	Additional overtime: Production and maintenance
2.11	Overtime worked by production and maintenance employees, by community setting (per cent)
2.12	Method of increasing overtime by shift length: production and maintenance employees (per cent)
2.13	Proportion of workforce who are contractors (per cent)

The impetus for this study came from the realization that, notwithstanding the lack of comprehensive data on new working arrangements, the rate of change in working time in the mining industry was not being matched by changes in other related parameters at the workplace, including those affecting occupational safety and health. The effects of extended shifts on workers’ exposure to ambient factors, such as noise, dust, heat, vibration and chemical agents appear to be largely unknown or, if they are known, heavily under-reported. The attendant issues of fatigue and long distance commuting have health, safety and social implications that warrant further study; not to mention the impact of fatigue on productivity.

The first step in any process of analysis is to obtain reliable basic data. This was the purpose of this comprehensive survey of the Australian mining industry. The results and the subsequent discussion of implications for occupational health and safety comprise this working paper

A survey design and pilot study were funded by ILO. The project was subsequently further supported by ILO and by the Departments of Minerals and Energy in New South Wales and Queensland and the Joint Coal Board Occupational Health and Safety Trust. The ILO is most grateful for these three contributions. The resulting jointly-funded project throws new light on the extent of working time in Australian mines and on the variety of shift and roster arrangements that are used. It also emphasizes the importance of undertaking future work on the health, safety, industrial relations and social issues that are affected by working arrangements.

The relevance and importance of the issue to the industry is underlined by the exceptional response that was achieved. Ninety per cent of Australian mines with more than 20 employees were contacted and 93% of them (180 mines) provided information. This is a strong response from an industry that has often been very sensitive about the issue of shift arrangements.

The ILO hopes that this study will be the precursor to similar work in other countries and lead to a concerted attempt to ensure that any additional occupational risks or hazards that are associated with extended working time and compressed shifts in mines will be identified and then eliminated or minimized without delay.

Oscar de Vries Reilingh
Director
Sectoral Activities Department

This report collates and analyses data on shift work arrangements operating in Australian coal and metalliferous mines collected by a national telephone survey conducted in late January and early February 2000. It is supported by qualitative and desktop research undertaken both as part of this project and as a consequence of other contracted research in the Australian mining industry by the authors. The survey was funded by the International Labour Office as a pilot study and was further supported by the Minerals and Energy Departments of the New South Wales and Queensland governments, and by the Joint Coal Board Occupational Health and Safety Trust.

The survey and study were undertaken as a result of the apparently increased prevalence of compressed and extended shifts across all Australian industries and the difficulties, noted in Australian Bureau of Statistics’ publications, in collecting reliable data on the nature extent of these changes. Available information indicated that the mining industry was the sector in which major alterations to working time arrangements such as average weekly hours of work, shift length and roster patterns, was most apparent. The mining industry is also characterized by the presence of local and overseas transnational firms. Moreover, it is particularly sensitive to international commodity markets.

The primary driver for the study was the apparent uniqueness of Australian shift work arrangements and the lack of data on their social, economic and OHS implications. A secondary, but significant driver was an attempt to draw together the possible implications of these changing arrangements. This project raises important issues about the future impact of these working time arrangements on the mining industry in Australia. By inference there may be international implications.

Section 1 contains the background to and context for the survey. This also includes a brief overview of the Australian mining industry – its place in the Australian economy, industrial relations/employment regulations and market trends – in order to put the trends into context and better understand them.

Section 2 presents the major findings from the national survey into shift work and rostering arrangement conducted in February 2000. The data are disaggregated by state, mining method and, as the Australian coal industry has some specific features, into "coal" and "non-coal" where appropriate.

Section 3 discusses the implications of these trends for working hours, occupational health and safety, family and community life, and industrial activity and trade unions.

Compressed shifts. Where an average of around 40 hours per week is worked in fewer days across a roster cycle. In this case time worked and time off are roughly symmetrical. For example, a typical compressed schedule may consist of four 12-hour shifts followed by four days off with an average of approximately 42 hours per week over a four week cycle. This means that over a 28 day cycle 16 days are worked instead of 20 under a typical eight hour roster. In some cases the compression cycle is longer (14 days worked followed by 7 days off) - typical of long distance commuting sites.

Extended shifts. Where time worked and time off are not symmetrical and more shifts are worked than are allocated in time off. In this case, instead of compressing the hours worked into fewer days, both longer days and longer average weeks are worked without being offset by time off. For example, instead of working 16 12-hour shifts over a 28 day cycle, employees may either extend individual shifts with overtime and/or work additional shifts.

Intensive schedule. A combination of long shifts (12 or more) combined with long overall weekly hours.

As the Commonwealth of Australia is a federation of states and territories the terms "Federal" and "Commonwealth" with respect to the national government are used interchangeably in the text.

The mining industry occupies an important position in the Australian economy. In 1996-97, the mining industry accounted for $22.6 billion (AUD) or 5% of Australia’s GDP. Over the last three decades the Australian mining industry has become one of the world’s largest with high levels of foreign investment. It is the nation’s leading individual export earning sector.

In 1997-98 Australia exported $9.6 billion worth of black coal (representing 11% of total merchandize exports), $6.3 billion of gold (7% of total merchandise exports), $3.8 billion of iron ore (4.3% of total merchandise exports) and $3.3 billion of bauxite/alumina (3.7% of total merchandise exports) (ABS, 2000). Its role in wealth generation and linkages with other key industries such as electricity generation, steel and transport make the mining industry an integral, powerful component of the Australian political economy.

Total employment in the mining industry at June 1997 was 55,713 (ABS, 2000: 477-79). Output growth has vastly outstripped employment growth, reflecting strong increases in capital and labour productivity. In coalmining, for instance, output has more than doubled since 1980 but employment has declined by just over a quarter from its peak in 1986 to 23,800 in December 1997 (Productivity Commission, 1999: 39 & 65). Large-scale retrenchments and restructuring of the industry occurred during the 1990s (Maher, 2000). The workforce is overwhelmingly male and older than the Australian average. This is more pronounced in the coal industry than in the metalliferous sector (ACIRRT, 1998). Although the direct employment generated by the mining industry is a relatively small part of the Australian labour market, it is a very important source of both direct and secondary employment in regional areas of the main mining states of Western Australia, Queensland and New South Wales.

The mining industry in Australia is still strongly unionized, but there are important differences between the metalliferous and coal sectors. Although the coal industry in New South Wales is still strongly unionized, the metalliferous sectors in Western Australia and, to a lesser extent, in Queensland have experienced declining levels of unionization.

There are also differences in trade union coverage and culture between the sectors. In metalliferous mining, the principal union is the Australian Workers’ Union (AWU), whereas the principal union in coal mining is the industrially powerful Mining Division of the Construction, Forestry and Mining Employees’ Union (CFMEU). This Division of the CFMEU, formerly known as the United Mine Workers Federation, is generally considered as more interventionist and socially active than the AWU. Other important unions in the mining industry are the Communication, Electrical and Plumbing Union (representing the small but strategically important electrical trades) and the Australian Manufacturing Workers’ Union (AMWU), representing engineering and mechanical trades. The CFMEU sometimes competes with the latter two unions for coverage of skilled-trade level workers.

In general, coal mining unions have established higher rates of pay and "job control" through extensive informal regulation, custom and industrial practice, than those in metalliferous mining (Hawke & Robertson 1999: 35; Wooden, et al. 1996).

The Australian mining industry, particularly the coal sector, has come under significant pressure during the 1990s. This is primarily due to the emergence of low-cost suppliers in Asia, such as Indonesia, increased competition in international markets and resultant pressures from domestic employers to restructure the industry and reduce costs associated with their operations. In addition, despite projections of continued growth in global demand, there has been a downward trend in international prices that is expected to continue. In black coal, for example, global demand is projected to grow by 2% per annum between 1997-98 and 2009-10 but real prices are projected to decrease by more than 2% per annum. Lower prices and competitive pressures have led to a squeeze on profit margins (Minerals Council of Australia 1999; Productivity Commission 1999; 41-42) and these tougher economic conditions have, in turn, placed pressure on regulatory arrangements and wage and employment conditions in the mining industry.

The primary source of employment regulation in Australia for most of this century has been the "awards" of federal and state industrial tribunals. Section 51 (xxxv) of the Australian Constitution gave power to the Commonwealth Government to legislate for "conciliation and arbitration for the prevention and settlement of industrial disputes extending beyond the limits of any one state" (Creighton & Stewart 1994: 56). As a consequence, and drawing on Australian and New Zealand colonial precedents, the Federation of the Australian colonies in 1901 saw the new Commonwealth Government commence to establish, at national level, the uniquely antipodean system of conciliation and arbitration tribunals. Similar tribunals operate at state level. These tribunals were delegated the authority to settle industrial disputes through adjudications placed in legally binding documents called "awards", usually covering an occupation or industry.

Australian industrial relations have undergone significant change over the last two decades, with the intensity of change escalating since the early 1990s. Not only have state and federal governments introduced changes to increase the type and range of bargaining processes and instruments, but profound labour market and industry changes have also underpinned a level of economic and workplace restructuring, the effects of which continue to reverberate across Australian society.

Although they remain an important source of industrial regulation, awards have been simplified and placed in the shadow of enterprise and individualized bargaining during the 1990s as part of a drive by policy-makers to encourage agreement-making and "workplace flexibility" (Gardner & Ronfeldt, 1996). Commencing in the late 1980s and early 1990s under a federal Labor Party administration, enterprise flexibility/award modernization clauses were placed in federal awards to allow enterprise-level variations of standard award provisions, including working hours. The Industrial Relations Legislation Amendment Act 1992 and the Industrial Relations Reform Act 1993 aimed "to ensure certified agreements are available as a real alternative to the award system and not reserved for exceptional circumstances" (Peter Cook, Federal Minister for Industrial Relations, cited by Hancock & Rawson, 1993).

For the first time, and under a Labor administration, provision was made for registering collective agreements to which unions were not a party. The new role of the tribunals was to facilitate enterprise agreements and maintain awards as a safety net underpinning enterprise bargaining.

The Workplace Relations Act (1996) introduced by the new conservative coalition Government which took office in 1996 built upon the foundations of this legislation by requiring awards to be simplified to just 20 allowable matters and further limited the powers of the tribunals. The original Bill (modelled on earlier New Zealand legislation) explicitly intended to replace collective arrangements with individual contracts. Although this intention was rejected by the Senate (the Australian upper house) and the Bill itself was substantially modified, the centrepiece of the new legislation was the provision to allow the registration of individual agreements that would over-ride awards. These are known as Australian Workplace Agreements (AWAs).

This legislation has seen a shift from an historical reliance on the industrial powers of the Constitution towards a reliance on the corporation powers. This has led to a re-emphasis on contract law applied to employment relations and has been accompanied by a range of regulations intended to inhibit collective action and, where possible, to encourage the individualization of the employment relationship. Over the same period, all state jurisdictions also introduced legislative changes to encourage enterprise rather than sectoral bargaining. Some states, for example Western Australia, have gone much further than others, such as New South Wales in promoting individual bargaining and the stripping or simplification of award conditions.

This labour market re-regulation, from national and sector level to enterprise and individual level, has had two primary aims. First, to move employees out of award regulation and into different bargaining streams (non-union collective, union collective and individual) and second, to restructure the residual award stream and restrict both the reach and range of issues with which it can deal (Campbell and Brosnan, 1997). This has allowed increased bargaining over a range of issues, including working hours, that were once determined within industry-wide and even community-wide frameworks. A secondary effect has, arguably, been the increased de-collectivization, individualization and resulting fragmentation of both wage and conditions outcomes (Heiler, Arsovska and Hall, 2000).

Where the primary regulatory framework governing the determination of wages and conditions (including hours) was formerly found in industrial awards, workplace relations, wages and conditions are now increasingly regulated under a combination of five different instruments:

• Federal and state legislation (industrial relations, occupational health and safety and anti-discrimination, corporation and contract law);

• The industrial awards of federal and state industrial tribunals;

• Enterprise agreements (union and non-union);

• Individual agreements (state and federal under industrial law);

• The (individual) employment contract (under contract and common law).

The relative importance, relationship and impact of these regulatory instruments have changed dramatically over the past decade. Federal and state legislation establishes the framework of rights and responsibilities for employers, unions and employees and minimum standards in some areas such as wages, some conditions, occupational health and safety and leave arrangements.

The importance of mining for economic development and the specific nature of the hazards associated with underground mining in particular had also resulted in some industry-specific legislation such as the former Coal Mining Act, 1925 (Queensland) and the Coal Mines Regulation Act, 1982 (New South Wales). These statutes focused primarily on the regulation of occupational health and safety in the coalmining industry. Even this legislation, however, is undergoing change as coal mines move away from prescriptive legislation and come under a more general duty of care obligation such as those introduced in Queensland under the Queensland Coal Mining Safety and Health Act (1999), which will come into force in September 2000 and some changes mooted in the NSW legislation covering coal, metalliferous and extractive industries. As will be noted later, the changes in this legislation will, for the first time, target the specific effects of working time arrangements in the form of requirements to pro-actively manage the effects of fatigue at a workplace level.

In mining, the bulk of employment regulations were determined by a combination of parent industry awards (e.g. the Coal Mining Industry (Production and Engineering) Consolidated Award 1997), occupational awards and secondary awards (which mostly follow the parent award) supplemented by site level over-award bargaining and informal regulations determined through custom and practice.

In fact, the mining industry, and especially the coal industry, has long been characterized by active workplace level bargaining. Local union lodges have always had considerable autonomy to develop and implement local arrangements that were usually, and significantly, above the award. However, there have also been sectoral differences in wage and working conditions outcomes between the coal and metalliferous sectors, with the coalmining sector, by and large, enjoying superior wages and conditions compared with the metalliferous sector. Historically this has largely been the result of a fluctuating combination of national and international commodity prices and local industrial bargaining power.

Over the past decade, however, the influence and impact of changes to industrial relations law have also emerged as significant drivers for change in the coal industry. The situation at the current time is still volatile. There is considerable bargaining over aspects of work organization (such as hours and shift work arrangements) that were previously viewed as heartland control issues for the union, but which have been undergoing sustained regulatory change. Indeed, one of the key areas of activity in mining agreements (as is also the case in other industries) has been to alter hours of work provisions and, in particular, to free up shift work arrangements (Heiler, 1998). This bargaining activity around hours of work has to be set against the context of weak industry or community-wide accepted standards about what constitutes safe or acceptable working time arrangements. Indeed, assumptions that working time standards contained within industrial awards would be preserved, despite increased and active bargaining, have proven to be, arguably, incorrect.

The regulation of working time in Australia has traditionally been undertaken almost entirely through the award system. Unlike many other OECD countries, Australia has no community-wide, legislative limitations on maximum hours of work, shift work or the length of the working day. Historically this was not seen as problematic since compliance with industry-wide award standards, and some state legislation which specified some isolated minimum standards, whilst not always consistent, did act as a de facto enforceable set of industrial standards governing working hours. Indeed, it was a model that, until the late 1980s and early 1990s, was reasonably successful in standardizing hours of work across and within industries.

At the beginning of the 1990s, the standard working time model codified in awards was for 35 – 40 ordinary hours per week, usually worked as an 8-hour day, with provision for any overtime to be paid at a penalty rate. Many industries also operated a 19-day month with one rostered day off (RDO) per month. Most awards contained detailed provisions associated with payment or compensation for additional hours (overtime) and non-standard hours (including shift work). Awards usually also contained provisions in relation to:

• The number and spread of ordinary working hours in each day;

• The number of ordinary hours per week (or sometimes a longer period such as a fortnight or month); and

• Special rates for work performed outside these hours (e.g. overtime, shift loadings, penalty rates).

Formerly, state and federal industrial tribunals had not been predisposed to expand the codification of "unusual" or extended working hours unless there were exceptional and compelling industry circumstances. Where shift work was permitted, such as in the manufacturing sector, the mining industry and emergency services requiring 24-hour operation, the award would usually define in detail the types of shift (starting/ finishing times of an afternoon shift or night shift) and the maximum length of the shift. It would also set out details governing overtime rates and hours and provide for the same number of ordinary hours as in ‘standard’ working time arrangements. There would also be some form of additional compensation for shift work usually by way of shift "loadings" that explicitly compensated for the personal and social costs that were deemed to be a result of non-standard hours. In other words, both in terms of the average length of the working week of around 40 hours and the structure of compensation that surrounded irregular and non-standard hours, well over 60% of the workforce, up until the late 1970s, appeared to fit within this standard profile of arrangements.

Over the past decade, however, the trend in Australia appears to have been to move away from this historically standardized working time model towards an increasingly fragmented and diversified one. In particular there has been a "hollowing-out" of the pattern of distribution and an increase in employees working either very long hours (in excess of 49 per week) and those working less than full-time hours. Average hours for full-time employees are now around 44 per week, an increase of about three hours per week in just eight years and a seeming ‘structural shift towards longer full-time hours across all industry groupings.

Figure 1.1 shows the decline in standard hours for full-time workers from 1978 to 1997. It reveals that less then 50% of full-time employees now work 40 hours or less per week, and that the most significant increase has been among full-time employees working very long hours of 49 or more per week.

This increase in average hours for full-time employees has been accompanied by a change in the distribution of hours, with the hollowing-out of standard hours across the labour market we referred to before clearly shown in figures 1.2 and 1.3 below. This hollowing has meant that large proportions of employees are clustered at one extreme or the other; either working far less than around 35 hours per week, or well in excess of 49 hours per week. There also appears to be a strong gender component to this trend, with males clustering around longer full-time hours and women clustering in part-time hours.

This shift in the distribution of hours in Australia appears to be most strongly associated with structural changes in the labour market. These include changes such as the high growth of part-time and casual employment, the workplace level restructuring that has seen significant employment reduction and an associated increase in work intensification and an increase in operating hours of organizations and companies which require more flexible and non-standard forms of labour (ACIRRT, 1999). This move away from standard hours of work (where the majority of the workforce worked 35-40 hours per week) has been most pronounced among males, but has also been significant among women. These trends are shown in Figures 1.2 - 1.4.

Figure 1.3 shows that these shifts in distribution are most pronounced for males. By 1998, for instance, 43% of all male employees were working in excess of 40 hours per week and almost one third (27.3%) were working over 49 hours per week.

Source: ABS: Labour Force Australia; Cat. no. 6203.0, August issues

This can be contrasted with trends for women. There was a sharp drop in the proportion working 40 hours per week in 1983 (from 23% to 12.5%) and a continuing increase in those working less than 30 hours per week.

Overall, the trend in hours of work since the late 1970s and especially the early 1980s has been one of steadily increasing hours for full-timers. At the same time there has been a hollowing-out of traditional standard hours across the labour market, most pronounced for males. These broad aggregate trends provide the context for understanding what has happened to average hours of work within the Australian mining industry.

As with other industries, the mining industry has experienced a significant increase in average weekly working hours since the early 1980s. However, the mining industry has not experienced the same hollowing-out and fragmentation in hours seen in most other industries, since the overwhelmingly form of employment in the industry is full-time. Instead, we have seen full-time hours steadily increase to the point where the mining industry has the longest hours profile of any industry in Australia. A number of different data sources illustrate these changes.

The increase in average weekly working hours across the industry has been steady since at least the early 1980s, when average hours across the industry were around 46 per week in 1983. By 1997, average hours had increased to 51.7 per week, longer than in any other industry. These trends are shown in figure 1.5. Not only does mining have the longest hours, it has also experienced the greatest increase in hours, accelerating between 1993 and 1997. Across the mining industry in 1997 the hours (51.7) consisted almost an extra full day (over seven hours) more than the all industry average (44.3).

However, these average increases mask a more complex sectoral distributional difference within the industry. Census data presented in Table 1.1 below shows that while both coal and metalliferous hours increased between 1991 and 1996, the proportion of persons working in excess of 41 per week in the metalliferous sector increased from 51.9% to 65.3% (13.4%), while in coal this proportion increased from 42.7% to 52.4% (9.7%).

The longer hours worked in the metalliferous sector illustrated by these aggregate trends were confirmed by our survey data outlined in Section 2 below. The results showed that, across the industry as a whole, over 70% of all employees worked in excess of 41 hours per week. In the metalliferous sector this figure was over 85% compared to around 50% of coal employees. This figure is even more stark when we compare the proportion of employees who work in excess of 49 hours per week. In coal this figure is around 4.2% while in metalliferous, more than 60% of employees work in excess of 49 hours per week. Thus we can see that whilst the mining industry as a whole has a very long hours regime, the metalliferous sector was primarily responsible in large part for these very long hours.

As a point of interest, the increase in hours of work across the Australian mining industry was also highlighted in international data contained in the ILO Yearbook of labour statistics that compared changes in hours between 1988 and 1997. Although average hours worked across countries are not necessarily directly comparable, the direction of increase is reliable. ILO data shows that while some countries appeared to experience a decline in average hours worked, Australia was one of the countries, along with the US, UK, Canada and Republic of Korea, that recorded a significant increase over this period.

Aggregate data on the extent, nature and detail of shift work arrangements across the Australian labour market have not been particularly robust. Part of the reason for this is that information about shift work arrangements and rosters is complex and difficult to collect; any one workplace may have several different roster systems and structures. The roster pattern can be stable or rotational and can be structured across a week or several months. We suspect that around 17% of all employees in Australia work some kind of shift arrangement (ABS, 1997), but the reliability of this figure can also be questioned. The reason is that the definition used for a shiftworker no longer reflects the large number of workers who now work outside of standard hours, many of whom may not identify as shiftworkers. Changes in the ordinary span of hours, increased operating of businesses, removal of loading and shift penalties, the annualization of hours and salaries all potentially place employees in situations where they can be called upon to commence work prior to 6am and finish work well after midnight (the traditional definition of a shiftworker). Traditionally the employees would have been classified as shiftworkers but the likelihood of them (information workers, service workers) labelling themselves as shiftworkers is lower than in traditional shift work areas (Heiler, 1998). In addition, traditional sources of data collected information about the details of the shift arrangements, especially with respect to shift length. This is partially because assumptions were made that shift length was stable at around 8 hours of ordinary time. It has taken some time for the data collection to catch up with the changes at a workplace level. The Australian Workplace Industrial Relations Survey (1995) did collect some limited information on shift length. This is useful to compare with the trends we detected in our survey. The AWIRS showed that in 1995, just over 10% of Australian workplaces had some or all employees who worked 12-hour shifts. While not precise, we can calculate from this data that in 1995 around one-third of mining workplaces had some or all of their employees working 12-hour shifts, with most of these in the non-coal sector (AWIRS 1995, unpublished data). This compares with the findings of our survey which revealed that over 50% of all mine sites nationally by 2000, and well over 70% in the metalliferous sector, worked 12-hour shifts. While not exact, we can detect a significant increase over this five year period.

In addition, we have seen that aggregate trends reveal both an increase in shift length and an increase in average weekly hours. Thus increasing sectors of the industry are working long daily hours combined with long weekly hours. In other words, the combination of our survey data and the aggregate hours data shows that the mining industry deviates from other industries in that a large proportion of the workforce are working not just long daily hours but long weekly hours as well. This means that the promise of compressed shift arrangements - of working fewer days and controlling hours within a standard range - is, by and large, not being delivered in the mining industry, and particularly in the metalliferous sector. The emergence of these intensive work schedules (QMC, 2000) can be briefly speculated upon.

Hours of work are now the subject of considerable bargaining at a workplace level across all industries, with both employers and employees seeking a level of flexibility that meets what each side asserts as its specific needs. While interest in changing hours of worked commenced in the 1980s, the focus on hours of work changes has been a central feature of both collective and individual agreements. Indeed, changes to hours of work provisions have been among the most common provisions in agreements across all bargaining streams since the early 1990s (ACIRRT, 2000). Figure 1.7 below shows that hours of work flexibility provisions are found in over 80% of enterprise level agreements, second only to provisions related to grievance procedures. This is true irrespective of the bargaining stream. Figure 1.7 shows the most common provisions in union and non-union collective agreements and individual workplace agreements (AWAs).

However, the extent to which these new forms of flexible working time arrangements genuinely benefit employees and the degree to which decentralized bargaining may be driving longer hours has been the subject of considerable debate and counterclaim (e.g. Buchanan and Bearfield, 1997; Campbell, 1997; Heiler, Arsovska and Hall, 2000; Wooden, 2000). While there is not necessarily a direct causal link between decentralized and individualized bargaining, there does appear to be a relationship between the kinds of provisions appearing in agreements, as is shown in Figure 1.8 below. Moreover, Wooden found in his analysis of the relationship between longer hours and form of bargaining that the high incidence of long hours at workplaces was most strongly related to the workplaces where individual contracts were in place (Wooden, 2000). Figure 1.8 shows how certain kinds of provisions in agreements are more likely to appear in non-union and individual agreements than in union agreements. These provisions are the ones most likely to be associated with changing or reducing compensation for longer or non-standard hours of work. These agreements include provisions such as increasing the ordinary days of the week and ordinary hours of the day which do not attract additional compensation, reducing overtime rates or time off in lieu provisions. While we may not be in a position to state categorically that there is a causal relationship between bargaining arrangements and trends in working hours, there is a relationship beginning to emerge.

What is becoming clear, however, is that the historical reliance on the award system, and the rapidity of working time changes have exposed an effective regulatory vacuum around working hours in Australia that is arguably leading to a level of unprecedented and unanticipated fragmentation and dispersion (Heiler, 1998a). As provisions in enterprise and individual agreements move away from the standards and structures contained within industry awards, they move also into uncharted and effectively unregulated waters where agreement about standards is generally absent. The emergence of new forms of bargaining-union, non-union and individual-has placed pressure on the traditional structures contained within awards for managing and setting the parameters for working time arrangements. These pressures exist strongly within the mining industry in Australia.

Mining industry employers have been at the forefront of moves to shift from awards to enterprise bargaining; especially from collective forms of employment regulation to individual agreements. This has occurred in a widespread way in the metalliferous sector, particularly in Western Australia where the Australian Workers’ Union has coverage. It has, however, met with greater resistance in the coal industry where the CFMEU (Mining Division) has primary coverage.

Until the late 1990s most enterprise agreements in the coal industry (as a result of a strong union bargaining position) did not entirely replace the award. Agreements generally contained few provisions that comprehensively dealt with issues traditionally covered by the award. A study of 27 coal industry agreements undertaken by ACIRRT in February 1997 showed that 70% of them contained provisions to modify shift arrangements in some way (ACIRRT, 1997). However, the majority of them did not contain provisions that dramatically lengthened shifts or reduced compensation for shift work.

These early agreements focused on increasing the ability to swap shifts, on variations to shift rates and on specifying the nature of the shifts (i.e.; fixed or rotating). Twelve-hour shifts had made an appearance in agreements by this time, with five of the 27 agreements analysed providing for these shifts. Significantly, one of these mines was the Vickery Mine at Gunnedah in NSW, owned by Coal and Allied (now Rio Tinto), where the introduction of these shifts became part of a bitter 12 month dispute. By and large, until the late 1990s in the coal industry, decentralized bargaining was occurring but was underpinned by a comprehensive award that protected basic conditions. The process of award simplification and declining union bargaining power in recent years suggests that more profound and structural changes are taking place.

As recently as June of this year, the Australian Industrial Relations Commission (AIRC) rejected a number of appeals by the CFMEU over the award simplification process that allowed employers unilaterally to extend ordinary hours of work. Foremost among the employers’ claims before the Commission was the "right" of employees to agree to longer hours or longer shifts. The AIRC determined that, in the interests of efficiency and productivity, "restrictions in the allocation of work" would be removed and the right of employers to impose 10-hour shifts without agreement, and to make application for 12-hour shifts would be upheld (Workforce, 2000). The Commission argued that these changes in working time arrangements were needed to "reduce labour costs and boost productivity" (Maher, 2000). In July of this year, after a long running and at times bitter dispute in the Hunter Valley at the Mount Thorley Mine with Coal and Allied (Rio Tinto), the CFMEU reluctantly signed off on an agreement that allowed management to determine start and finish times without negotiation, to introduce 12-hour shifts with a "hot-seat" changeover and to allocate additional overtime as required. Coal and Allied’s Managing Director stated that: "It gives us everything we have been seeking by way of flexibility, productivity and management control" (Long, 2000).

The move to individual employment contracts has been most successful in the metalliferous sector in Western Australia, accentuated by a high level of contract employment across the industry. Inroads are continuing to be made in the coal industry but the rate of growth has been far slower. From the early to mid-1990s, particular transnational mining corporations operating at multiple sites embarked on a drive to lessen the role of third parties by offering large pay increases conditional on the acceptance of increased flexibility and, in some instances, individual contracts and/or reclassification to staff positions. Major disputes have flared periodically, either at the instigation of unions trying to check the spread of individual employment contracts (e.g. Weipa [1995] ) or at the instigation of managers, some of whom it has been claimed, view:

The passage of the Workplace Relations Act (1996) and the introduction of AWAs:

Changes to longstanding working time arrangements have been an integral part of these new arrangements across all industries. In general, it has been assumed that what is flexible and efficient for business is also, virtually by definition, progressive for employees. This has particularly been the case with respect to the emergence of non-standard compressed and extended working time arrangements such as those within the mining industry.

It is to an examination of these merging trends that we now turn.

The population for the survey was all operational coal and metalliferous mines of more than 20 employees. Preparation and processing plants in mining and quarrying operations were excluded for cost reasons. The survey frame was drawn from the 1999 Australian general mining year book and the 1999 Australian coal year book (AMS, 1999). All mines in these yearbooks with more than 20 employees and classed as operational were included. The primary unit of analysis was the mine site, rather than individual employees. Information on the roster that most employees usually worked in production and maintenance at that site was collected. This was to enable the most common arrangements for the majority of employees to be determined. This approach was adopted, rather than attempting to describe all site rosters, since there are a vast number of possible individual rosters, although only a small number of employees might work any particular roster.

The protocol was developed by the researchers and it was administered by a commercial survey company. An initial telephone screening was conducted to eliminate non-operational mines and to confirm contact details. The senior person with knowledge of shift roster arrangements was identified. A total of 217 mines were identified and 180 useable surveys completed. There were 13 refusals, seven confirmed as closed and a further 14 that could not be contacted. Excluding closures, this represented an overall response rate of 86%. From the 196 mines able to be contacted, a 93% response rate was secured. This meant that overall there was a 7% refusal rate. This can be interpreted as a very strong response from an industry that has been highly sensitive about the issue of shift arrangements.

Mining activities are unevenly distributed across Australia. As table 2.1 below shows, the majority of mines responding to the survey were in three states, Western Australia (32%), New South Wales (31%) and Queensland (23%). These three states

Table 2.1 Distribution of responding mines by state (in order of number of mines)

Source: ACIRRT ILO Survey, February 2000 (Population, all mines, n=180 responses).

When grouped into main commodity types, the results show that black coal (40%) and gold mining (29%) provide the largest proportion of sites (69% of respondents; table 2.2 below).

Table 2.2 Major products produced by mines (all respondents)

Source: ACIRRT ILO Survey, February 2000 (Population, all mines, n=180 responses).

Coal, which represents the greatest number of sites, is almost exclusively concentrated on the eastern seaboard, with Queensland and New South Wales together containing 90% of all coal mines.

Overall, open cut operations are the most common (49%), followed by underground (40%) and mixed operations 11% (table 2.3).

Table 2.3 Mining Methods, by State (per cent)

Of the three major mining states (WA, NSW, Qld.), New South Wales is the only state to have a higher proportion of underground mining. This reflects the high proportion of underground coal mines in NSW. Overall, Queensland is more broadly comparable to Western Australia in its mixture of mining operations than it is to New South Wales, despite the high proportion of coal mines in the two eastern seaboard states.

The time of operation gives a good indication of potential rostering arrangements. As mines move towards continuous operation it appears the incentive is greater to move towards compressed arrangements (12-hour shifts), particularly if sites are remote or isolated.

The survey results show that there are significant differences between the coal and metalliferous sectors and among the three major mining states. As can be seen in figure 2.1, 70% of all Australian mines operate on a seven day a week basis, with Western Australian mines most likely to operate on a continuous basis (93%). Queensland is just above the national average at 75% with New South Wales far less likely at 38%. The greater percentage of seven-day operations in Western Australia is not purely a function of remoteness or isolation, since almost one-third of these mines are in settled, mixed communities. Coalmines are also far less likely to operate on a continuous seven day per week basis than are non-coal mines.

The difference between the states and sectors for days of operation is sustained for daily hours of operation, although to a lesser degree. Overall, 86% of mines operate on a 24-hour basis. Of those mines not working on a 7-day/24-hour basis, the largest proportion are in the coal industry, where 53% of mines work less than seven days a week and 21% work less than a twenty four hour day (figure 2.2 below).

Western Australia again stands out as the most likely state for mines to operate 24 hours per day, while Queensland conforms to the national average. Interestingly, while only around 38% of New South Wales mines operate for seven days per week, double this number operate 24 hours a day. This suggests that daily continuous daily operation does not necessarily imply seven days per week operation, (at least in New South Wales), whereas it is the general rule in Western Australia and Queensland.

The community setting of the mine was also significant for type and times of operations. Overall, 62% of mines are located in a settled mixed community, with 16% in relatively isolated mining towns and a further 22% operating on a "fly-in/fly-out" basis. As figure 2.3 shows, the tendency is for coal mines to be in a community setting; a large proportion of Western Australian mines to be ‘fly-in/fly-out’ operations; and the location of New South Wales mines to be in a community setting. In Queensland the majority of sites (56%) are located in a mixed community setting with 23% in isolated communities. Western Australia and Queensland, the two largest states in area, have the most long distance commuting sites; with Western Australia (49%) and Queensland (21%) using fly-in/fly-out operations. The Northern Territory, large and sparsely populated like Western Australia, has six out of its nine mines operating as open-cut operations, with the largest mines, also fly-in/fly-out, operated from isolated mining towns.

Derived from: ACIRRT ILO Survey, February 2000. (‘all mines" n=180); (‘coal’ n=77); (‘non-coal’ n=103) (‘Qld.’ n=41) (‘NSW’ n=55) (‘WA’ n=57)

Not surprisingly there appears to be a relationship between mine location and mine operating hours. Table 2.4 below indicates that almost all mines operating for less than seven days per week (48 of 49, col.1) and all those operating less than 24 hours per day (column 3) are within settled/mixed communities. In Queensland, only one mine operates less than seven days per week and less than 24 hours a day (coal), while only three mines operate less than 24 hours a day. All four are within a settled community, with the three mines operating less than 24 hours a day also operating with 12-hour shifts.

The potential complexity and multiplicity of rostering arrangements meant that the survey focused on collecting descriptive data about the main rosters at each site surveyed. Sites were asked to specify standard shift length, rotation pattern, number of consecutive minimum and maximum shifts, ordinary average hours of work and overtime practices.

Respondents were asked to provide identical information about both production employees and maintenance employees. As table 2.5 indicates, there is a high proportion of 12-hour shifts in mining. There are also some differences between production employees and maintenance employees, with maintenance workers less likely to work 12-hour shifts.

Table 2.5 Standard shift lengths for production and maintenance employees:
all respondents all mines (per cent)

However, as figures 2.4 and 2.5 show, there are significant sectoral and state differences. In particular there is a much higher proportion of 12-hour shifts in the non-coal area and in Western Australia. In W.A. we can see that some form of 12-hour shift arrangement predominates in the production area (93%), while in N.S.W. at 16%, and in the coal industry as a whole (25.4%), the opposite is true. On average, Queensland (at 50%) lies between the Western Australian and New South Wales extremes, with Queensland virtually a mirror image of NSW with respect to the distribution of shift patterns. In Queensland, 85% of sites have a standard shift length in excess of eight hours; in NSW this figure is 39%. Western Australia, in comparison, has no sites with eight-hour shifts as their most common production shift.

Although coal mines account for over 50% of the number of Queensland mines, there is no difference between the frequency of 12-hour shifts in Queensland coal or metalliferous mines. The difference between Queensland and Western Australia is therefore not a function of the high proportion of coalmines in Queensland. The coal industry difference that appears in the data really reflects the New South Wales coal industry. It is therefore not a function of differences between coal and non-coal in general, but between the NSW coal industry (particularly the underground coal sector) and other sectors.

Interestingly, of the ten Queensland mines that stated a desire to change roster arrangements, all were coalmines. The desired direction of change (identified in the qualitative responses) was towards increased shift length or hot seat provision in the production areas, and increased shift length, hot seat provision and rotating shifts for maintenance.

A comparison of figure 2.4 with figure 2.5 shows that 12-hour shifts are less common in maintenance than in production. Note in particular that 12-hour maintenance shifts in WA are more than three times as common than they are in Eastern coalmines. Again note that while 50% of coalmine sites have maintenance shifts of 12-hours or more, another 32% have shifts of more than eight hours duration.

Derived from: ACIRRT ILO Survey, February 2000. (‘Maintenance’ n=150).

There is an argument that the extent of compressed shifts is primarily a function of the isolation of a particular site. However, if we remove the long distance commuting (LDC) sites from the analysis, we can see that this is not the case. Figure 2.6 shows that 12-hour sites in production fall only 2 percentage points in Western Australia and around 7% in Queensland. We can therefore conclude that compressed shift arrangements are not simply or predominantly a function of the isolation of the mines.

One question that can justifiably be asked is whether shift length is at least partly determined by mining method. In other words, do we find that open-cut mines are more likely to have compressed shifts than are underground mines, where (arguably) the environment is less suited to long working days? Table 2.6 below reveals that compressed shifts are far more common in open-cut environments. 67% of all open-cut mines have their main shift length at 12 hours, while in the underground sector this figure is 40%. Still, it is significant that 40% of underground mines have a standard 12-hour shift length in the production area. The figure is a little less for maintenance, both in open-cut and underground mines.

Standard shift length is only one factor associated with how shifts are worked. The ways shifts are compressed (or extended) are important components of roster design and are highly significant for OHS, family and social life. The major identified differences in shift and roster patterns were related to sector and to state. As figures 2.7 and 2.8 show, the trend towards multiple consecutive shifts is most marked in Western Australia and in the non-coal sector.

In the coal sector, only 1-2% of sites work more than seven shifts in a row, with between 86% (production) and 94% (maintenance) working 5-6 days in a row. In the non-coal sector, 28% work 14 shifts in a row; in Western Australia this rises to over one-third (34%). The figure for WA is significant since we saw previously that 93% of sites in production work 12-hour shifts. For Queensland, half the sites reported standard 12-hour shifts. The difference between Queensland and Western Australia is not attributable to the high proportion of fly-in/fly-out operations in WA. Even eliminating Queensland’s LDC sites, WA has proportionally more sites operating 12-hour shifts and with a greater number of consecutive shifts. A "coal industry effect" is noticed even in Queensland, where about one-third (6 of 20) of its coalmines operate a three or four day roster.

As figure 2.8 shows, the position is similar for maintenance workers.

Figure 2.9 shows that almost one-third of sites that work on 12-hour shifts also work them for 14 days in a row. In WA this figure is over one-third and in Qld it is 23%. No NSW 12-hour sites work any more than four days in a row. In Queensland over 50% of sites with 12-hour shifts work them five days or more in a row, compared to 81% of WA sites. Even when we excluded fly-in/fly-out operations the relative differences remained. It would appear, therefore, that shift patterns are more than just a function of the relative isolation of the mine, although this does have some effect.

As we saw earlier, the relative isolation of the mine was not strongly associated with the likelihood of the mine to have compressed 12-hour shifts. However, the intensity of the compression appears to be influenced by the relative isolation of the mine. Figure 2.10 shows that when LDC sites are removed from the analysis, the percentage of rosters where the shifts are worked in very long runs of consecutive shifts (i.e. 14 or more) falls markedly, especially in Western Australia. It is significant, however, that even at sites where employees are commuting on a daily basis, there are still many sites where the shifts are heavily compressed. For instance, almost 40% of sites in WA have shifts that operate seven days or more in a row, and in Queensland this figure is around 35%.

There were some reported differences between maintenance and production workers in the type of roster cycle, with maintenance workers almost 10% less likely to work rotating shifts (47%) than production workers (57%), (Table 2.7).

We know from other data that average hours of work have apparently increased in Australia in the last 15 years. We also know that the mining industry has the longest ordinary and total hours profile of all industries. The survey confirms this, showing that over 60% of all sites have average ordinary working hours well in excess of 40 per week.

Figures 2.11 and 2.12 show differences between coal and other mines, with coal mines again less likely to report very long ordinary hours. Again, in terms of state differences, Western Australia is above the all mines Australian average. Queensland has a longer ordinary hours regime than NSW, but not as long as WA.

Again, community setting was an important factor in hours of work, with longer ordinary weekly hours associated with the relative isolation of the mine site (Table 2.8).

Figure 2.13 shows the trends for ordinary average hours of work on sites where 12-hour shifts are the most common arrangement (all mines). It shows that longer average ordinary hours appear to be associated more strongly with 12-hour shifts. Just over 50% of 12-hour sites have ordinary average hours in excess of 49 per week, compared with an average of 31% for all sites. In principle, there should be no difference between the average weekly hours of work at 12-hour shift sites compared with any other shift/roster arrangement. This is because 12-hour shifts are meant to be worked within a compressed hours regime, not an extended hours regime (see Wallace, 1998). The longer average hours associated with the 12-hour sites suggests that many of these arrangements are being worked within an extended hours regime rather than in a compressed regime.

Overall, approximately 70% of both production and maintenance employees worked overtime on either a regular or occasional basis. (Table 2.9). This overtime is in addition to any mandatory overtime already absorbed into standard hours as a result of an industrial award, agreement or individual contract.

The inclusion of mandatory overtime is a common feature where 12-hour shift arrangements have been introduced. This may explain why WA sites are less likely to be working additional overtime (figure 2.14). Figure 2.14 also shows that coal mines are much more likely to work regular overtime than the non-coal sector. This is because their ordinary average hours are lower and these employees are compensated separately for overtime worked.

Of the 70% of employees who did work overtime, the large majority worked less than eight hours per week, with over one-third working 4-8 hours. Again, there are differences between the production and maintenance areas (Table 2.10) with maintenance workers more likely to work fewer overtime hours (i.e. under 4 hours per week) but production workers more likely to be working substantial