In Search of the Magic Pudding

Compulsory Workers' Compensation: Worker Right or Unnecessary Restriction?

R. Ian McEwin

'Workers' Compensation is a liability neither in tort nor in contract. ... The parties may choose whether they will enter into the relationship; but if they do the employer's liability for, and the worker's and his dependant's right to, compensation are legal consequences which are independent of and cannot be controlled by their agreement.'

Dixon J. in Mynott & Orrs v Barnard2

Introduction

Workers' compensation legislation makes employers liable for government-prescribed disability benefits if injured or killed at work, or on the way to and from work. But do workers gain from legislatively provided 'rights' to compensation? Increased legal 'rights' to compensation do not yield the benefits claimed by many. For many workers, increased legal accident 'compensation rights' were largely paid for by reductions in previously negotiated compensation for taking risks. Legislatively imposed 'compensation rights' effectively restrict a worker's ability to negotiate alternate, preferred, compensation packages because the 'compensation right' is inalienable. As a result, workers may suffer welfare losses. Worker welfare would be improved by permitting workers to opt out of the workers' compensation system and negotiate desired accident compensation arrangements directly with employers.

Workers' compensation legislation also makes employers buy insurance to cover their liability on terms and conditions determined by government. Increasingly, in Australia, employers have been forced to buy their compulsory worker disability insurance from government insurance monopolies. But has increased government control led to disability insurance being provided at the lowest possible price, consistent with worker preferences? Has industrial safety been improved?

This paper argues that, consistent with experience in other industries, excessive regulation and government monopolisation of the supply of disability insurance has increased insurance costs and restricted the availability of desired insurance policies. Considerable improvements in worker welfare and resource allocation would be achieved by reducing the role of government and returning decisions about industrial safety and disability insurance to the workplace level.

Background---Common Law Compensation for Industrial Accidents

Most societies have developed rules (or tort law) to ensure injurers compensated victims. However, tort law did not assume its modern form in the English-speaking common law countries until the 19th century both because intentional wrongs were seldom litigated and because non-contractual, non-domestic accidents were relatively rare. Industrialisation and the advent of the railways dramatically increased the numbers of accident victims.

Little is known about the extent of employer liability for workplace accidents before 1837. In that year (which preceded the industrialisation which took place in the latter half of the 19th century in Britain), the English High Court, in its first reported case dealing with an employee suing his employer for injuries negligently caused at work, decided that an employer was not liable for injuries negligently caused by a co-worker.3

By the middle of the century, English common law courts had established rules under which a worker might sue an employer for compensation. These principles differed from those governing non-contractual relations, being grounded on the premise that employer liability was based on an employer's promise, either express or implied, to compensate injured workers, which was in turn based on the peculiarities of the workplace situation. To recover damages, a worker had to prove that the accident resulted from an employer's lack of reasonable care (or negligence) in selecting workers and/or maintaining safe plant and equipment. Reasonable care was defined in terms of the amount of care a reasonable man would take in the circumstances. In other words, employers were made liable in circumstances where they were likely to have more information about the risks involved than workers and where they were in a better position to prevent injury loss (where injury was likely through defective machinery, etc.). In the remaining cases, workers were expected to have more knowledge about the risks involved and a greater ability to prevent loss. A higher standard of care was required in more hazardous employment, such as mining. Court concern with allocating costs to those best able to prevent accidents suggests industrial safety was perceived as the main goal of the common law, not compensation.

Employers could defeat most claims by relying on three common law defences. First, contributory negligence which ruled out any recovery if the worker contributed, however slightly, to the accident. Second, assumption-of-risk which, in accordance with a general belief that courts should not be involved in voluntary agreements, assumed that workers willingly accepted the possibility of injury. Third, common employment which absolved employers from injuries caused by other workers 'in common employment'. The assumption of risk and common law defences effectively meant that employees were less likely to recover (non-contracted) compensation from an employer, than an injured stranger. Injured railway passengers were able to recover, but not injured railway workers. This supposed lack of equal access to the law was to prove to be an important argument against the common law by no-fault reformers. However, presumed inequality before the law ignored the fact that where bargaining is possible, risk compensation can be negotiated before an accident occurs. In particular, workplace accident compensation was provided via bargaining between employees and employers. Additional wages were paid to workers in dangerous occupations (e.g. coal miners) to enable them to make their own accident insurance arrangements. Sometimes, compensation was negotiated between workers and employers through more generous sick leave and accident insurance schemes compared to non-dangerous occupations.

Early Legislative Changes to Workplace Common Law Liability

The latter part of the 19th century saw considerable debate, both inside and outside the courts, on the extent of an employer's liability for industrial accidents. Focusing on the doctrine of common employment, trade unions argued that workers should be put on the same basis as non-workers. By placing a greater and more certain financial load on employers, not only would workers be compensated but employers would have more incentive to take safety precautions. Employers argued, on the other hand, that increased employer liability would have serious consequences for industry. The Employer's Liability Act was passed in England in 1880, and served as a model for similar legislation in Australia and the United States. The 1880 Act displaced common law liability rules in a number of specific instances but did not affect the assumption of risk and contributory negligence rules. Injured workers were put in the same position as injured strangers for injuries negligently caused by coworkers---the defence of common employment was restricted to situations involving defects in plant and equipment and for situations where the victim was working under the direction of another worker. However, the statutory liability was not binding as employers and workers were allowed to 'contract out' of the statute's provisions.

Reformers, shocked that so many workers subsequently contracted out of the Act's obvious 'benefits', introduced a large number of Bills into the British Parliament in the last two decades of the 19th century. Most Bills proposed the abolition of contracting out and the various common law defences as well as increases in the maximum compensation payable. Australian parliaments closely followed British developments.

Public debate centred around the principle of freedom of contract---which provided the basis for the common law doctrine of assumption of risk. Reformers argued that workers were often in weak bargaining positions and had little choice but to accept dangerous work. Unscrupulous employers, it was argued, did not bear the costs of accidents and so had few safety incentives. Labour unions and some employers argued that more humane employers were put at a competitive disadvantage because they paid more compensation and spent more on safety. Reflecting the times, no empirical evidence was produced to support these propositions nor was the extent of negotiated risk compensation examined. Instead, it was assumed that giving additional legal rights automatically benefited workers. Instead, worker rights to negotiate desired compensation was reduced and replaced by government fiat.

Accident Compensation in Australia Before Workers' Compensation

Unfortunately, little data are available on how injured workers fared in Australia before the introduction of workers' compensation. But the available evidence suggests that while compensation may have been difficult to obtain through the common law, extensive accident compensation arrangements existed independently of the courts. Prior to the introduction of workers' compensation, injured workers in Australia were largely reliant on the charity of employers and others; the benefits from membership of mutual societies such as friendly societies, trade unions, health societies; employer-sponsored sickness and accident schemes; government provided medical hospital care, free dispensaries and other institutions; and sometimes the common law.

To a large extent workers' compensation replaced these former, mainly private arrangements.4 For example, union membership often included disability benefits. Nairn (1973, p.423) lists accident pay, sick pay, unemployment pay as common union benefits. Cass (1983, p.21) points out that two out of six new unions examined (and formed between 1860 and 1890) included accident benefits in their rules. The Miners Protection League, formed in 1861, provided medical benefits. This union was to become the Amalgamated Miners' Association in 1890. Accident pay provided through that union amounted to 2 per week for 12 months and then 10 shillings per week thereafter.

At the turn of the century about one-third of the Australian population was covered by friendly societies in some way. Coverage was concentrated in manual trades with high accident rates e.g. miners. It has been estimated that about 80 to 90 per cent of manual workers were members (see Inglis, 1880, p.2). Australian friendly societies were based on their British equivalents and reflected mid-Victorian ideals such as self-help, work and success. Relations with the trade unions were close, and both offered members social facilities not available elsewhere. Many friendly society meetings were held in hotels, although a considerable number of friendly societies were also associated with temperance associations. Middle-class men played a large part in the development and running of friendly societies, not only reflecting the tendency for the middle classes to preach to the working classes, but also because the societies were often seen as a stepping-stone to political and business success.

Nobbs (1983) suggests a number of reasons for the decline of friendly societies towards the end of the 19th century, including: an increase in publicly provided social activities by governments such as parks, libraries, museums; a decline in interest in issues such as patriotism, nationalism and temperance which had been the main reason for the formation of many societies; the growth of life insurance which competed directly with societies for death or retirement benefits; and perhaps most importantly, the increased intervention by governments into welfare areas that the societies had pioneered.

Life assurance offices also provided death and endowment policies. 360,000 policies were in operation in 1901 (out of about 1.3 million employed males) . Most policies were written by Australian companies who operated with minimal government control in a highly competitive environment. Nobbs (1983, p.465) noted that:

'In the 1870s and 1880s the Australian insurance market had no peer. Nowhere else had policyholders the same advantages nor was life cover per head of population greater. The changes in life assurance practice were to influence insurance business throughout the world.'

State government also provided social welfare benefits indirectly. Hospitals were heavily subsidised, fees accounted for only about 10-15 per cent of total hospital revenues, with half of the remainder coming from government grants and the rest from private charity and fund raising. However, self-help was emphasised. Apart from non-employment-related benefits, employees could obtain compensation from their employers, either on a charitable basis, through company sickness and accident schemes.5

A number of income maintenance programs were introduced at the beginning of this century. Aged and invalid pensions were introduced in 1901 in Victoria. New South Wales introduced an age pension in 1901 and disability pensions in 1907, the Commonwealth in 1908 and 1910 respectively. The schemes were non-contributory, with flat-rate means-tested benefits. Disability pensions covered permanent incapacity, irrespective of the source of that incapacity.

Unemployment insurance was considered by the Fisher Government in 1911. The Cook Government in 1913 announced a National Insurance Scheme, to be funded on a contributory basis, 'covering sickness, accidents, maternity, widowhood and unemployment'.6 Nothing came of the scheme, partly because of friendly society opposition. Some government intervention was designed to encourage the use of self-help mechanisms. For example, the NSW Government introduced an Act in 1908 subsidising friendly societies for long-term sickness cases.

Introduction of Workers' Compensation in Australia

The British Workers' Compensation Act of 1897 served as the model for the introduction of similar legislation in Australia. Workers' Compensation Acts were introduced by South Australia in 1900, Queensland in 1905, New South Wales and Tasmania in 1910, South Australia in 1911 and Victoria in 1914.

British historians have emphasised the importance of the passage of workers' compensation legislation in terms of a gain in working class rights, which did not exist under the common law. Bartrip (1986, p.5), for example, has said:

'The symbolic importance of the Act was in conferring a working class right---comparable with the right to vote or to form trade unions---which, if it did not remove the longstanding disadvantages of the negligently injured worker relative to those injured outside the workplace, promised to circumvent them.'

The Act has been interpreted similarly in Australia, being seen as part of a 'new liberalism' in Australia at the beginning of the 20th century (see Crowley, 1974, pp.260-311, and Fitzpatrick, 1969, pp.241-68). Changing attitudes about the economic role of the State led to a growing belief that individuals had 'social rights', to be ensured by government, including the right to a minimum income. As a result, in addition to workers' compensation, old age and invalid pensions were introduced and minimum wages laid down. However, in contrast to invalid pensions, which were designed to provide income support for a limited number of disabled, workers' compensation was '. . . based on the principle that workers as a class had a collective right, a just claim, to compensation for work related injury from their employers' (Cass, 1983, p.80).

While the common law was seen as anti-worker in both Britain and the United States, the fact that unions opposed workers' compensation, preferring instead the abolition of employer defences (and so keeping compensation in the hands of common law judges) was never explained. Merritt (1982, pp.76, 81) argues that workers often used the common law, successfully, to improve employment conditions:

'Since it is the worker who stands to lose by legal enforcement of contract entered with minimal formality and fleshed out by unequal implied terms, the nineteenth century workers' attempted insistence on agreements concluded only after full discussion and acceptance of express provisions demonstrated a keen grasp of the way in which the law could be used to protect their interests within labour relations . . . Even under the body of law surrounding the Masters and Servants Act, the rights of workers unable to work through illness or injury were frequently at issue. While today the situation has been resolved by legislation providing compensation for injury at work and by clauses in industrial awards, 19th century workers attempted to establish their rights through contract.'

The extensive negotiations about risk and disability arrangements contradict romantic claims made by some legal historians that stress a lack of worker power in the employment relationship and which argue that workers were forced to accept risk. Workers' compensation legislation effectively replaced employer/employee bargaining and imposed conditions on the employment contract with respect to safety and accident compensation.

Historians and economic historians have largely avoided examining the introduction of workers' compensation in Australia. The notable exception is Cass (1982) who examined the introduction of workers' compensation in New South Wales. Cass argued that trade unions were instrumental in the introduction of workers' compensation in New South Wales through the lobbying of parliamentarians and pressure on Labor governments.7 Conservative opposition, she argued, was weakened due to divisions between employers who opposed the legislation and (mainly British) insurance companies who supported it. Insurance companies strongly supported the legislation. It gave them the opportunity to expand their business at the expense of mutual self-help groups such as friendly societies. The Chamber of Manufacturers in New South Wales, fearing high insurance premiums, responded to the legislation by establishing their own insurance company---now called Manufacturers' Mutual Insurance Ltd.8

Unlike the United States (see Croyle, 1978), employers groups opposed the legislation, arguing that the insurance premiums would impose unreasonable costs on industry, especially for small business. Concern was felt about the effects of workers' compensation on industry--- which prompted the Liberal Party to propose a contributory National Insurance Scheme. In the end the view prevailed, as noted in the High Court, that '. . . the industry itself should be taxed with an obligation to indemnify the sufferer for what was an accident causing damage'.9

Apart from Victoria, no State initially required employers to insure their liability, although a number of politicians had spoken in favour of a State insurance scheme (see Cass, 1983). Many employers insured their liability in any case.10 Within a few decades, all employers were required to buy insurance from approved insurance companies. New South Wales introduced compulsory insurance in 1926 and at the same time established a (competitive) government insurance office. Queensland introduced compulsory insurance in 1916 and also established a State-owned monopoly.11 Western Australia introduced compulsory insurance in 1925, Tasmania in 1927, the Australian Capital Territory and the Northern Territory in 1931 and South Australia in 1932.

By simply imposing liability on employers for certain amounts in the event of worker injury, workers' compensation is distinguishable from other forms of early social intervention in Australia like old age and invalid pensions. While those other interventions provided benefits from general taxation revenues, workers' compensation made no claims on the public purse.

A major difference between workers' compensation in Australia and the United States was that, in the latter, workers' compensation was explicitly seen as being traded for restrictions on common law rights. It was believed that both employers and employees gained from the legislation. However, common law rights were retained in Australia. As mentioned previously, the vigorous defence of common law rights by trade unions in Australia does not support the claim that the rights were of negligible importance, nor that judges were anti-working class. However, Australian evidence on the actual numbers of employees compensated through the common law system either directly or indirectly (through non-litigated settlements) does not exist.

While there seemed to be a general consensus among the various political parties in Australia that some form of industrial disability scheme was called for, there was considerable disagreement about levels of benefit and who should pay. There were four unsuccessful attempts to introduce the legislation in New South Wales (the first in 1889) and at least six in Victoria. The principal concern was the effect of increased costs on industry. The Liberal Party proposed a contributory National Insurance Scheme. In general, all States adopted similar legislation which provided for payment of compensation by an employer to a worker for personal injury by accident 'arising out of and in the course of employment'. By 1913, only South Australia had extended the provisions to include industrial disease. Even there, the industrial diseases only covered poisoning (anthrax, lead, mercury, phosphorus or arsenic). All States, apart from Queensland and Tasmania, covered workers travelling to and from work.

Employers were exempted from liability for injuries which did not disable the worker for a minimum period. In 1913 the minimum disabling period was two weeks in New South Wales, three days in Queensland and, for the remainder, the first week of injury if disabled for less than two weeks. Maximum limits were imposed on employer liability under the legislation. For example, the maximum total liability for incapacity in 1913 was 200 in New South Wales. Employers were not liable for injuries resulting from the wilful misconduct of the worker. Minimum periods before which lump-sum payments could be substituted for future weekly payments also applied. Most States adopted a six months minimum period. Not all workers were covered initially. Most schemes were restricted to manual labour in dangerous industries. In New South Wales, for example, the Act applied to manual labour employed in railways, factories, mines, quarries, wharves, vessels, engineering or building work and other proclaimed employment. Miners were covered under similar provisions in the Miners Relief Act, which preceded workers' compensation.

Workers' compensation simply imposed strict liability on employers, for government determined benefits, for injuries suffered by workers in dangerous occupations. Nothing else changed. Worker rights to sue negligent employers for full compensation were retained. All States prohibited contracting out of liability under the Act, although schemes (without worker contributions) could be substituted, provided benefits were not less than provided under the legislation. Substitute schemes could not be made compulsory. In South Australia, Western Australia and Tasmania, schemes were permitted which gave additional benefits in return for worker contributions. These schemes had to be approved by a majority of workers as well as the government. Over the years, workers' compensation schemes have undergone substantial changes. Benefit levels have increased and coverage extended to occupational disease and the journey to and from work. In addition to the earlier income maintenance and safety objectives, rehabilitation is now also considered to be of major importance.

Initially, workers' compensation disputes were handled by the courts, for example in magistrate or district courts. Gradually the role of the courts was reduced by legislation, firstly by placing greater emphasis on arbitration but secondly through the establishment of separate workers' compensation commissions whose role was not only to settle disputes but also to fix premium rates, license insurers and generally regulate the workers' compensation system in each State.12

Unlike the United States, workers in both Britain and Australia retained the right to sue a negligent employer through the courts. Access to the common law has also been made easier by legislative means, for example the defence of common employment has been repealed. However, in the last five years common law rights have been restricted in some States.

The evolution of the current system has been characterised by increasing government involvement over the terms and conditions under which insurance is supplied. At the time compulsory insurance was introduced, there was widespread political agreement that employers needed to be protected from the 'excesses' of insurers, given a captive insurance market. This concern was understandable given the uncompetitive nature of the insurance industry at that time: it was largely controlled by a cartel of British insurers. With hindsight, a preferable action would have been the promotion of insurer competition. Workers' compensation insurers were licensed and subject to prudential and price controls. Government insurers, competing with private insurers, also were set up in some States: Queensland, for example, introduced a government monopoly in 1916. Private insurers were required to lodge returns concerning their business. However, the lodgement of returns was not strictly policed, nor their accuracy checked (Cooney, S 6.2), which accounts for the poor quality of data and the resultant inability of official enquiries adequately to diagnose the causes of the financial problems in the system.

The Current Scene

Workers' compensation consumes considerable resources in Australia. In 1982-83 (the latest year for which aggregate Australian data can be compiled from official sources), workers' compensation premiums exceeded $1.3 billion, or $205 per worker (in 1986-87 prices). Some figures from the Victorian WorkCare system provide more recent data. In 1986-87, reported claims in Victoria totalled 204,064, or one claim per 7.4 employed persons. Payments by WorkCare amounted to $333m, and outstanding liabilities of WorkCare rose from just under $717m at 30 June 1986 to $2.64 billion at 30 June 1987.

Large increases in premiums in the early 1980s led to widespread employer discontent, particularly in the manufacturing and construction industries. Between 1981 and 1983, the average annual growth in the hourly cost of workers' compensation in Australia was 49.2 per cent. The premiums of some firms in Victoria increased by 200-300 per cent between 1981 and 1982. State performances differed. For example, workers' compensation premiums per worker rose in New South Wales from $67.58 in 1978-79 to $224.90 in 1982-83 (in constant 1986-87 dollars). Equivalent figures for Victoria were $70.46 and $221.08, for Queensland $53.51 and $132.99, and for the ACT $31.52 and $29.41.

Insurers argued that rapidly rising premiums, regulated by government, were not increasing fast enough to cover the more rapidly escalating claims. State governments saw high workers' compensation premiums adversely affecting the competitiveness of their exporting and import competing industries. As a result, seven official enquiries have been conducted into workers' compensation in Australia in the last five years. But none was able to explain, satisfactorily, why costs were blowing out. Nor were the enquiries able to account properly for considerable differences in workers' compensation premiums and claims costs between States. For example, in 1984-85, claims incurred as a percentage of wages varied from 0.82 per cent in the ACT to 2.76 per cent in New South Wales, to 3.23 per cent in Victoria (Craigie, Cumpston and Sams, 1986, p.26).

While premiums have in general been rising, the number of claims made have remained reasonably constant. For example, in New South Wales the amount of compensation paid increased from $199.01m in 1977-78 to $387.20m in 1984-85 (in 1986-87 prices). Claims declined from 244,329 to 223,669 in the same period. The main reason for the cost increases seems to be increased income maintenance payments per claim, due to increased benefit levels and the associated incentives to remain 'on compo'. These explanations are consistent with overseas evidence. In the United States, for example, a number of studies have found that both the frequency of claims and the duration of time spent away from work rise with increased workers' compensation benefits (Worrall and Butler, 1985). Similar problems have been encountered with the comprehensive accident compensation scheme in New Zealand (Business Roundtable, 1987).

Why do workers' compensation costs vary so much across States? Possible explanations include differences in: labour markets (resulting in differences in willingness to bear risk or make claims); industrial composition---States have different proportions of less safe industries; the cost of living; the level of compensation benefits; the extent of government safety regulation; the extent and form of government insurer regulation; and the degree of competition in workers' compensation insurance. All of these factors need to be taken into account in evaluating the relative efficiency of each system. So far this has not been done.

As the official enquiries did not identify the principal causes of increasing claims, they were unable to assess whether past government regulation (of safety or insurance) had contributed to the problems. Nevertheless, more government intervention was recommended. Major changes have been made to workers' compensation arrangements in Victoria (WorkCare), New South Wales (WorkCover), South Australia (WorkCover) and, to a lesser extent, the Northern Territory (WorkHealth) since 1985. The new schemes restrict common law rights and reduce the role of private insurers.

Past enquiries into workers' compensation, reflecting, perhaps, the domination of the workers' compensation system by lawyers, have placed too much emphasis on making incremental changes to existing arrangements and have failed to make comparisons of alternative systems of compensation and safety promotion. Hampered by legal blinkers, accident compensation has been seen in terms of finding 'deep pockets' to pay for increased benefits. Scant regard has been paid either to the option of facilitating greater competition in the provision of insurance, or to the promotion of more enterprise-oriented negotiations about industrial safety and accident compensation arrangements. Current government-imposed workers' compensation schemes involve both compulsory employer liability for government-determined benefits and the regulated supply of compulsory liability insurance. The amounts and the conditions under which benefits are paid have become more precisely specified by government, and the terms and conditions under which insurance is supplied have become more constrained. Alternative sources of compensation through the common law have been restricted or abolished, and insurance supply has been concentrated in the hands of State government monopolies.

Some Basic Workers' Compensation Economics

Although everyone would like to eliminate accidents (defined here to include disease), the reality is that accidents are costly to avoid. Prevention only can be achieved by expending scarce resources or forgoing risky activities altogether. People knowingly risk personal injury because the benefits of engaging in risky activities are judged to be worth the costs. As with other activities, people will typically choose not to work in a risky job unless there are compensatory monetary or non-monetary benefits, relative to the remuneration from less risky jobs. Put bluntly, the Evel Knievels of this world should not be banned, nor should they be eligible for subsidised insurance. Given the risks, Evel fans should pay prices for tickets sufficient to pay the extraordinary income or the high costs of insurance, or both.

Public concern about the appropriate level of safety is diminished if each person is fully informed about the risks involved and if the full benefits and costs of engaging in the risky activity accrue to that person or enterprise. Safety precautions will tend to be taken by individuals and employers so long as the additional safety level expected is perceived as worth the additional costs. The preferred level of safety and the preferred form of insurance differ between individuals because people have different attitudes to risk, form of compensation and the costs of safety precautions. A necessary consequence of this is that government policy should be directed at promoting the mechanisms through which people achieve their desired safety and compensation outcomes, including a desired range of insurance services.

In practice, accidents are jointly caused, in the sense that their likelihood depends on the preventative actions of both parties (or co-workers). Workers were compensated for accidents before the enactment of workers' compensation legislation. To attract workers to risky occupations, employers had to pay higher wages, additional sick leave, etc. Workers bought disability insurance through friendly societies and life insurance companies. Workers' compensation simply replaced, in whole or in part, risk coverage and compensation arrangements privately negotiated between workers and employers.

The notion that workers and employers can negotiate and receive risk and remuneration packages irrespective of legislatively determined compensation is crucial to a proper understanding of a number of workers' compensation issues. While employers will tend to have a comparative advantage in some aspects of risk determination, workers also affect outcomes, and, at the enterprise level, are capable of negotiating preferred mixes of insurance, safety conditions, work practices and wages. Much of the recent concern with labour on-costs (reflected in the 1986 Report on the Costs of Workers' Compensation in Australia by the Federal Advisory Committee on Prices and Incomes) is misplaced when it is realised that workers' compensation replaces other elements of workers' package of pay, leave, safety and other conditions.

Another misconception relates to the abolition of common law negligence actions. Risk-related remuneration can, and typically would, account for the likelihood of a successful tort action; and if access to tort actions is abolished, negotiations would seek an adjustment upwards in other forms of compensation for risk.

Why then have employers been concerned with workers' compensation rates? Why have some of them lobbied for the socialisation of what should be competitive insurance arrangements? One important reason has been the wide variation in premiums for similar firms in different States, particularly in manufacturing. This led, it was argued, to firms in States with high workers' compensation premiums suffering a competitive disadvantage. Another problem is simply that governments imposed obligations on firms, for example, to cover the first five days compensation, and discouraged innovation in design of insurance packages capable of offering workers higher wages, but different rights to accident pay and leave. Differences in insurer competitiveness and government actions restricting premiums (and so cross-subsidising some industries either at the expense of other industries now or in the future) also have given legitimate cause for employer concern.

Table 1

Assessment of policy options

Goals of Public Policy Towards Industrial Accidents

The goals of accident policy are non-controversial. Most official enquiries argue that government safety and compensation policies should:

• promote efficient levels of safety which means well-informed decisions by workers about participating in risky activities, and cost-justified levels of care and safety expenditure by employers and workers; and
• provide adequate compensation in the event of accidents, with coverage against loss of income and reasonable medical, hospital and rehabilitation expenses.

These sub-goals are consistent with the goals mentioned by a number of official enquiries into accident compensation, in particular the US National Commission Report and the Report of the Committee of Enquiry into the Victorian Workers' Compensation System. However, these goals should be pursued only so long as the benefits gained are likely to be worth the costs. In the interest of safety as well as equity, the actuarial costs of accidents should be covered by insurance arrangements struck between those benefiting from the risky activity and those who are in the best position to reduce accident risk.

Alternative Policy Options

The broad range of policy options by which governments can promote safety and accident victim compensation are:

• changing the common law, which shifts the responsibility for payment of accident losses through liability rules;
• safety regulation;
• compulsory self- (or first-party) insurance, with varying degrees of control over insurance supply;
• compulsory employer liability insurance, again with varying degrees of control over insurance supply;
• social security.

Each policy approach has advantages and disadvantages. Table 1 provides a comparison of the way each option contributes to safety and compensation objectives.

The common law

Legal responsibility for accident losses can lie where they fall (no liability or victim liability), or be shifted by common law liability rules. Two liability rules predominate in the common law, negligence and strict liability. Under strict liability, employers are liable for personal injury

loss, irrespective of their contribution to an accident. Negligence liability, on the other hand, only makes careless employers liable for loss. In the last century, negligence liability was the norm. But over the course of this century, both through judicial and legislative changes, strict liability has become more important.

The issue of whether employers or workers should be liable for accident losses is, surprisingly to some, not the key to safety and compensation arrangements. If competitive and efficient disability insurance is available to workers either through group policies or employer policies, then different groups, firms and perhaps industries will opt for different liability arrangements. As discussed previously, informed workers will demand extra remuneration to work in jobs with higher risk of personal injury. On the one hand, if employers are not liable, the extra remuneration will tend to be taken as wages and used to buy the amount of insurance they want. On the other hand, if employers are liable, the extra labour cost will be met by a lower wage than in a no-liability situation. Further, if employers are obliged to compensate injured workers in an artificial or non-competitive manner through arbitrary WorkCare-determined rules, which involve extra costs, this will be regarded as part of the risk-remuneration package and so wages often will have to be correspondingly lower, compared to a competitive insurance situation. An employer negligence rule would be an intermediate situation. Workers will demand less compensation for risk under the negligence rule than under the no-liability situation, to reflect the increased likelihood of obtaining compensation.

Workers, then, may ultimately end up with similar levels of expected real remuneration, irrespective of the liability rule in force. However, the level of accident insurance taken out will differ because courts might make employers pay more compensation to an injured worker than would follow from the amount of insurance that workers would buy. Employers take the full level of accident costs into account, and have an incentive to take care in order to reduce the amount of insurance premiums and/or wages they have to pay. It turns out that workers and employers have similar incentives to take care under either liability situation if there are competitive insurance agreements on safety prevention, and if wages and work practices are negotiated and monitored by each side.13

While workplace liability is not the key to workers' risk compensation arrangements, nor safety, it does matter to the costs of administering the system, since without employer liability no adjudication or administrative costs are involved. So long as workers have insurance, no liability rules may often be preferred, partly because they overcome the problem of employers with limited assets under employer liability. Undoubtedly, the reason strict employer liability was preferred at the turn of the century was the absence of a social security net and, perhaps, misperceptions regarding first-party disability arrangements at that time. Other explanations revolve around the advantages to unions and insurers of breaking up the competitive self-help organisations like friendly societies. In the absence of social security, employers were seen as convenient sources of funds. Politically, the approach was acceptable because of the widespread and mistaken belief that employers did not have to compensate for accidents in the workplace. They did, but it was typically in the form of higher wages than would otherwise have been the case.

Safety regulation

Safety regulation attempts to influence safety expenditures and care by controlling dangerous activities directly, for example by specifying safety standards for processes and products and regulating their use, subject to financial penalties. Both care and activity levels may be affected by such arrangements.

Due to a lack of information, safety regulators face high inspection and compliance costs as each plant, product and so forth must be examined. Also, rapid changes in technology can alter the optimum level of safety and form of safety regulation. Safety regulation involves, then, a relatively expensive administrative process compared to tort liability; the latter deals only with a small number of actual accident cases. And of course safety regulation does not provide compensation.

Compulsory insurance

Insurance, unlike any of the other instruments, can both promote safety and provide compensation. More importantly, because the trade-off between safety and coverage is addressed directly, improved safety and compensation are likely to result. Governments can make disability insurance compulsory in two ways: by making (groups of) workers buy insurance directly or by making employers buy disability insurance on the workers' behalf.

Social security

Medicare and social security pay medical expenses and provide a basic level of income maintenance to all. As they are financed from general taxation revenues, they do not impose the costs of accidents on the sources of risky activities nor do they provide any safety incentives. As a result, more risky activities are, effectively, subsidised by taxpayers.

Which Option?

Social security effectively relieves industry of an obligation to meet a base level of costs associated with work accidents but does not provide any safety incentives. And safety regulation, by itself, does not provide compensation. This line of reasoning suggests that the major policy design options facing governments are two-fold: either compulsory worker (first-party) disability insurance or compulsory employer liability insurance; and, guidelines and regulations regarding the supply of compulsory insurance. In deciding between compulsory first-party and employer liability insurance, three criteria can be identified. First, the extent to which the alternative approaches permit tailoring of policies to enterprise and worker preferences. Second, the efficiency with which each type of insurance can be supplied. Third, the extent to which each contributes to efficient levels of safety.

If the level of compulsory insurance were set at 100 per cent of loss, and, assuming workers and employers had identical information and could monitor agreements perfectly at zero cost, there would be no substantive difference between the outcomes under employer and employee liability. However, if minimal obligatory levels of employer liability insurance are set (say equal to the social security level), then first-party insurance would be preferred because workers could then 'top-up' to the desired extent by purchasing additional first-party coverage.

The second and third criteria both relate to the efficiency of supply of insurance. The main issues here are whether first-party or liability coverage is cheaper and the relative efficiency of competitive insurance markets versus a State monopoly. No simple answer is possible to the first. In large companies, or large unions and/or worker associations, collection of premiums and risk assessment may be less costly, allowing cheaper group insurance for workers. For many small companies, self-insurance might be no more expensive and yet provide more suitable terms. The diversity of insurance requirements suggests that there are no grounds for imposing one form of insurance. There is no reason, for example, why an enterprise should not have a mixture of policies. An employer could take out a liability policy for those workers who want it, a union could take out group cover on behalf of its members, and other workers could take out their own insurance contracts.

The more important issue is whether insurance should be supplied through competitive private insurance markets or by State monopoly. Despite its importance, the recent official enquiries in Australia have failed adequately to deal with this issue. One important reason for this is the ideological belief that workers' compensation is part of social security and so should not be in the hands of private enterprise. The Victorian Trades Hall Council, for example, in its submission to the recent Victorian enquiry into workers' compensation said:

'The Trades Hall Council is committed in principle to the concept of a single State instrumentality for the handling and processing of workers' compensation claims. It is philosophically abhorrent to have control of a social service in the hands of private enterprise' (quoted in Cooney, 1984, S 5.10).

This attitude is naive. If we genuinely are concerned with worker welfare, then we should be concerned primarily with ensuring that maximum benefits are paid from limited compensation resources, not with philosophical stances that are likely to disadvantage workers.

Data are not available to give definitive answers on the relative merits of competitive versus monopoly insurance supply. However, the overwhelming empirical evidence from other industries is that inadequate competition (e.g. in the financial sector, of which the insurance industry is part) leads to higher prices and to less incentive to control internal costs, creating a strong presumption in favour of competition. More careful analysis of the problems caused by excessive regulation has led to a trend towards more competitive workers' compensation arrangements in the United States (Williams, 1986). The Victorian Workers' Compensation Enquiry asserted (S 5.12) that a government monopoly had advantages over private insurers because the profit element was removed and because State monopoly was excused from State taxation. In a national context these claims are quite false. First, even if a government monopoly does not have to make a profit, there is an opportunity cost of using the funds involved. Lower workers' compensation premiums due to the absence of a profit element almost certainly will be more than offset by the losses resulting from using those funds in a more productive use. The same point applies to the taxation argument. The reduction of taxation revenues means other public spending opportunities are lost or additional taxation is required elsewhere. The social benefits of these forgone activities must be compared, from a public viewpoint, with any advantages gained from lower workers' compensation premiums.

The only other serious argument in favour of a monopoly is the advantage of size. No recent empirical evidence is available on scale economies in workers' compensation insurance. However, the large number of private insurers in nearly all States in the 1970s suggests scale economies are reached very quickly. Competition promotes efficient risk classification. Insurers have an incentive to attract low-risk customers away from competitors by offering lower rates or innovative policies. This process also serves to inform customers about the actual risk they face and gives them an economic incentive to reduce risk. State monopoly insurers have insufficient incentives to promote efficient levels of safety by engaging in more finely tuned merit-rating and dissemination of information about accident risks.

Moreover, public insurers may find themselves under political pressure to subsidise particular groups in particular high-risk industries who have the most to gain. And they also have a resulting tendency to suppress access to information on claims, risks and difficulties, largely out of a desire to hide information on cross-subsidies. This has happened already in New South Wales where the concept of 'community rating' has been introduced to '. . . overcome economic circumstances as in the case of trade threatening manufacturing' (NSW Government Discussion Paper, 1986, p.46).

In Victoria, it has been remarked that the origins of WorkCare had much to do with some influential employers managing to persuade governments to 'socialise' their risks, rather than look to more competitive and fair solutions appropriate to their firms and industry, or to better risk monitoring at the enterprise level. The ability of governments and powerful employers to use the workers' compensation system for other ends is a cause for concern, particularly as subsidies to relatively unsafe industries and production technologies mean that their level of activity is greater than otherwise. As a result of this artificial risk-spreading, the number of accidents is probably greater than otherwise because relatively unsafe industries are encouraged to expand (as premiums are less than actuarial costs) and relatively safe industries are discouraged (as premiums are greater than actuarial cost).

Various surveys of costs of provision of insurance have noted that the ratio of benefits paid to premiums received is generally higher for public insurance companies, and have then proceeded to assert that a government monopoly scheme is superior to a competitive industry. This line of argument is subject to serious flaws. At worst, a high ratio of benefits paid to premiums received would be due to underfunding or to the overpayment of claims. More subtle reasons for differences include differences in the quality of service offered (such as supplying too little safety advice, inadequate claims investigation, inadequate monitoring of claimants). The check to administrative waste is, in any case, competitive entry, and this is prevented with a government monopoly.

Monopoly insurers also have insufficient incentives to minimise costs. When overall funding is largely unconstrained, the government monopolist has options to request price increases to cover inefficiencies on a cost-plus basis, or even to fund deficits from general revenue. Often, there are incentives to promote programs publicly to satisfy political constraints, irrespective of worth. Monopolies have insufficient incentives to reveal or solve problems in the system that conflict with managerial goals. For example, solutions to excessive claims will be biased towards more employment of staff rather than improved monitoring of safety, assessment of risk and adjustment of premiums. By contrast, in a competitive industry, with the threat of new entry, private firms have a profit incentive to provide new products more attuned to industry and worker needs and to reduce costs below those of competitors.

Conclusion

We need efficient and competitive insurance arrangements that give workers the cover they want at least cost, and which also provide incentives for appropriate safety prevention by workers and employers and which facilitate proper monitoring. This is unlikely to be achieved by the current practice of precluding workers and employers from negotiating their own workplace safety and risk-compensating arrangements. Excessive government regulation of insurers and State insurance monopolies are part of the cause of higher costs and a poor safety record in Australia. Legislative intervention in workplace disability and insurance arrangements, rather than enhancing worker 'rights' and welfare, has unnecessarily restricted efficient levels of disability coverage and insurance supply. Current WorkCare style policies are, we suggest, likely to lead to more not less claims for accidents and to worse insurance outcomes than if competition was fostered in the private sector. The preliminary evidence from WorkCare in Victoria and from the New Zealand experience suggests that a government monopoly does not improve matters---on the contrary, it is far worse than allowing normal but competitive insurance processes to prevail, with government setting some basic ground rules, for example in relation to compulsory insurance and insurance company prudential requirements.

Endnotes

1. This paper draws on 'Safety, Disability and Compensation', in J. Freebairn, M. Porter and C. Walsh (1988).

2. (1939) 62 C.L.R. 68 at 91 (High Court of Australia).

3. Priestly v Fowler (1837) LJ. EX. 42.

4. McEwin (forthcoming 1988b) provides evidence that coal miners were paid compensation for taking risk. This occurred despite the existence of Industrial Tribunals which said they would not pay 'blood money'.

5. Cass (1983, p.87) cites evidence that the Colonial Sugar Company (CSR) provided a better workers' compensation scheme than one proposed in 1889, and that large companies such as Burns Philp had always provided paid sick leave.

6. Ministerial Statement CPP 1913, Vol.III, p.137, quoted in Kewley (1973, p.141).

7. As does Sutcliffe (1967, p.70) and Fitzpatrick (1968, p.110).

8. Pursell has noted that insurers were also 'the only significant group which supported the extension of the early Employers Liability Acts' (1964, p.321).

9. McGuire v The Union Steamship Company of New Zealand, (1920) 27 C.L.R., p.583.

10. This contrasted with the situation in the United States where many workers' compensation statutes made employers buy insurance. States that did not make insurance compulsory removed common law defences (employee negligence and assumption of risk) for non-insured employers (see Fishback, 1986).

11. Perhaps surprisingly, insurance was not made compulsory in Britain until 1946, when workers' compensation was replaced (the coal industry was made to buy insurance from 1934).

12. The introduction of specialised workers' compensation boards comprised of compensation experts did not necessarily lead to a lessening of legal influence. For an account of the re-legalising of workers' compensation in the United States see Nonet (1969).

13. These points are taken up in more detail in McEwin (forthcoming 1988a).

References

Bartrip, P.A., 'The Rise and Fall of Workmen's Compensation', Manuscript, 1986, Centre for Socio-legal Studies, Oxford University.

Cass, G., Workers' Benefit or Employers' Burden: Workers' Compensation in New South Wales 1880-1926, 1983, Industrial Relations Research Centre, Sydney.

Coase, R., 'The Problem of Social Cost', Journal of Law and Economics, 1960, 3, 1-44.

Cooney, B.C., 'Report of the Committee of Enquiry into the Victorian Workers' Compensation System', Government Printer, Melbourne, 1984.

Craigie, R., Cumpston, R and D. Sams, 'Accident Compensation Reform', Australian Economic Review, 3rd Quarter, 1986, 9-30.

Crowley, E., A New History of Australia, 1974, Heinemann, Melbourne.

Croyle, J.L., 'Industrial Accident Liability Policy of the Early Twentieth Century', Journal of Legal Studies, June, 1978, 7(2), 279-97.

Epstein, R.A., 'The Historical Origins and Economic Structure of Workers' Compensation Law', Georgia Law Review, Summer 1982, 16(4), 775-819.

Fishback, P.V., 'Workplace safety During the Progressive Era: Fatal Accidents in Bituminous Goal Mining, 1912-1923', Explorations in Economic History, 1986, 23, 269-81.

Fitzpatrick, B., The British Empire in Australia---An Economic History, 1834-1939, 1969, Macmillan.

Freebairn, J.W., Porter, M.G. and Walsh, C., Spending and Taxing II: Taking Stock, 1988, Allen and Unwin, Sydney.

Inglis, J., Our Australian Cousins, 1880, Macmillan, London.

Joskow, P., 'Cartels, Competition and Regulation in the Property-Liability Insurance Industry', The Bell Journal, Autumn, 1973.

Kewley, T.H., Social Security in Australia 1900-1972, 1973, Sydney University Press.

Mayerst D. and Smith, C.W., 'Contractual Provisions, Organisational Structure and Conflict Control in Insurance Markets', Journal of Business, 1981, 55, 407-34.

McEwin, R.I., 'Workplace Accident Compensation Reform: A Reappraisal', Australian Economic Review, (forthcoming 2nd Quarter, 1988a).

'Risk Compensation and Coal Mining: Some Empirical Evidence', Centre of Policy Studies, (forthcoming 1988b).

Merritt, A., 'The Historical Role of Law in the Regulation of Employment---Abstentionist or Interventionist?, Australian Journal of Law and Society,1982, 1(1), 56-86.

National Commission on State Workmen's Compensation Laws, Compendium on Workmen's Compensation, U.S. Government Printing Office, Washington, 1973.

Nairn, N.B., Civilising Capitalism, The Labour Movement in NSW, 1870-1900, 1973, ANU Press, Canberra.

New South Wales Government, 'NSW Workers' Compensation Scheme: Options for Reform', Discussion Paper, 1986, NSW Government Printer.

New Zealand Business Roundtable, 'A Submission to the Law Commission', July 1987.

Nobbs, R., 'The Development of Friendly Societies and Life Assurance in Australia', Journal of the Institute of Actuaries, 1983, 110,457-66.

Nonet, P., Administrative Justice, 1969.

Palmer, G., Compensation for Incapacity, 1979, Oxford University Press.

Pursell, G., 'The Development of Non-Life Insurance in Australia', Unpublished ANU Thesis, 1964, Canberra.

---'Rate Control and Government Competition in Australian Non-Life Insurance', The Journal of Risk and Insurance, 1967, 237-54.

Rabin, R.L., 'The Historical Development of the Fault Principle: A Re-interpretation', Georgia Law Review, 1981, 927-61.

Rea, S A., 'Private Disability Insurance and Public Welfare Programs', Journal of Public Finance, 1981, 36, 84ff.

---Disability Insurance and Public Policy, University of Toronto Press, 1981.

Ruser, J.W., 'Workers' Compensation Insurance, Experience-Rating, and Occupational Injuries', Rand Journal of Economics, Winter, 1985,16(4), 487-503.

Williams, C.A., 'Workers Compensation Insurance Rates: Their Determination and Regulation', in J. Chelius (ed.), Current Issues in Workers' Compensation, 1986, W.E. Upjohn Institute for Employment Research, Kalamazoo, Michigan, 209-35.

Woodhouse, O., Report of the Royal Commission of Inquiry: Compensation for Personal Injury in New Zealand, 1967, Government Printer, Wellington.

---Report of the National Committee of Inquiry into Compensation and Rehabilitation in Australia, 1974, AGPS, Canberra.

Worrall, J.D. and Butler, R.J., 'Benefits and Claims Duration', in J.D. Worrall and R.J. Butler, Workers' Compensation Benefits: Adequacy, Equity and Efficiency, 1985, ILR Press, New York, 55-70.