The Relation Between Function and Form in the Main Federal Industrial Tribunals

Introduction

There is a rich literature on the Commonwealth Court of Conciliation and Arbitration (CCCA) and its successor tribunals. We know a good deal of the origins of industrial tribunal regulation and something of its place in our social and economic history (see Isaac and Macintyre, 2004; Macintyre and Mitchell, 1989). There is a significant body of literature also on the law of the tribunals, treating their jurisdictions and powers and criticizing decisions of the day (a landmark is Kirby and Creighton, 2004; see also Creighton and Stewart, 2010). Yet attempts to understand the character of the procedures deployed by the tribunals, in particular in the exercise of their arbitral functions, have arguably been limited. They have not sufficiently isolated the relation between function and form. This creates potential problems for understanding the nature of the tribunals, making comparisons with other tribunals, assessing recent attempts at tribunal redesign and understanding the contribution of the tribunals to the rule of law in Australian industrial relations.

In an effort to highlight and begin to ameliorate these problems, this article draws on process jurisprudence and reinterprets the adjudicative functions of the main federal industrial tribunals. In the first part, some important conceptual accounts are reviewed: first, Bennett's (1994) class-based critique of the role of courts and tribunals in industrial relations; and, second, those of Dabscheck (1980, 1981, 1995, 2004) and Frazer (2006) on the tribunals as ‘regulators’. In the second part, Fuller's analysis of legal processes, in particular his idealization of an adjudicative process, is outlined and suggested as a foundation for understanding what is distinctive and important about the work of the tribunals (on Fuller's work generally, see Winston, 2001).

The third part will sketch the relation between the nature of the tasks allocated by the legislature to the tribunals and the structures and procedures of those tribunals, with particular reference to the Australian Industrial Relations Commission (AIRC) and Fair Work Australia (FWA). The aim of this part is to lay the groundwork for an analysis that draws out how these bodies have adapted key elements of the adjudicative ideal. Of special interest are the distinctive ways that the tribunals have mediated the participation of the industrial parties, the approach to complicated tasks such as wage-fixing and the need for expedition. The focus is on procedural design and avenues for research into the tribunals’ contributions to the rule of law in the industrial context. This should, in turn, lead to a better understanding of the continuities, discontinuities and dilemmas represented in the modern tribunal, FWA.

Some preliminary notes are necessary. First, the following discussion refers to the ‘main federal tribunals’. By this is meant, successively: the CCCA from 1905; the Commonwealth (later the Australian) Conciliation and Arbitration Commission (CAC) from 1956; the AIRC from 1988; and FWA from 2009. While there has been important continuity in terms of the functions, procedures, personnel and main litigants in these bodies, their succession has not been entirely lineal. A break occurred after the High Court of Australia's decision in the Boilermakers case, ¹ which triggered the federal government to separate out the CCCA's judicial/enforcement and conciliatory/arbitral functions into a court and commission, respectively. The creation of FWA by the Fair Work Act 2009 (‘FW Act’) represents another significant break (Stewart, 2011). ²

Second, each of the tribunals has discharged multiple regulatory functions, and none more than FWA, which lend themselves to distinct procedures (for a brief description of the range and evolution of functions, see Acton, 2011). The focus of this article, however, is on those decision-making functions that appear to be significantly adjudicative in nature, initiated by applications and determined by orders and a formal decision or similar. It should also not be forgotten that adjudicative processes in the tribunals have generally been preceded by informal conciliation conferences. This is important, not least because it is at this stage that most industrial matters are resolved (see MacDermott and Riley, 2011; Mourell and Cameron, 2009; Provis, 1997). But conciliation is a different kind of legal process and, at some risk, it is not examined here.

Third, this article does not examine how the High Court's interpretations of the Constitution have impacted on the tribunals, nor the supervisory and enforcement functions of courts in relation to arbitration, though plainly they exercise very important functions in industrial law through a distinctly adjudicative process (see Creighton and Stewart, 2010: paras 6.47–62). ³ Industrial tribunals in the States, special tribunals and other regulatory agencies and commissions also cannot be examined in this space, even though a number of these have deployed interesting modes of adjudication and have, at times, assumed as much or more importance in workplace regulation than the main federal tribunals (Creighton and Stewart, 2010: paras 6.63–70; Romeyn, 1986).

Conceptual Accounts of the Main Federal Tribunals

In this part, three theoretically informed accounts of the main federal tribunals are examined: those of Laura Bennett, Braham Dabscheck and Andrew Frazer. All three have made important contributions to our understanding of these tribunals, but arguably they do not isolate the relation between function and form or tell us what is important about the tribunals’ modes of adjudication and why.

Bennett (1994) offers a powerful socio-legal critique of the different roles of courts and industrial tribunals in the making of Australian labour law. She focuses on conflict in the creation of these fora and the struggles for influence within them, between them and with governments. Of particular relevance is her account of how the structures and personnel of courts and tribunals determine what interests enter into their decision-making processes.

Bennett's critique of the role of courts is a radical one: ‘The courts have played an important role in subverting labourist legislation and promoting a disciplinary approach to arbitration’ (1994: 64). This is said to stem from the application of English common law doctrines, and because the judiciary and lawyers have been drawn disproportionately from affluent and conservative upper- and middle-class ranks: ‘The social relations of the judiciary, their professional socialisation and selection all serve to reinforce anti-union values and biases’ (1994: 66). She supports this with analysis of the destructive application of common law interim injunctions to employee industrial action, and judicial support for bans clauses in awards (Bennett, 1994: ch. 3).

The shortcomings of the judiciary, perceived and actual, are said to have informed the development of the main federal tribunals and other tribunals in the employment field (Bennett, 1994: 69). Bennett distinguishes two tribunal forms: judicial and contextualized. Judicial tribunals function like courts and ‘tend to deal with individuals as discrete units’ (Bennett, 1994: 107). The problems of common law courts’ interventions in labour relations, in terms of procedure and more fundamentally ideology, tend to be replicated. On the other hand, contextualized tribunals, into which category the main federal tribunals after 1956 may fit, are ‘integrated into the industrial relations system’ (Bennett, 1994: 108). They are required by legislation to pursue policy objectives, rather than, like courts, to simply determine rights. Contextualized tribunals are designed to facilitate collective representation and be more responsive to industrial dynamics (Bennett, 1994: 109), which may bring them into conflict with superior civil courts (Bennett, 1994: 74–84, citing a number of episodes of conflict between the courts and the main federal tribunals).

A key dynamic in the development of the main federal tribunals has been the fact that tribunal members and litigants are engaged in a continuing relationship (Bennett, 1994: 100). Constant interaction is described as a structural constraint on tribunal practices and decision-making, because to maintain their relevance and ‘institutional standing’, the tribunals must meet the needs of the industrial parties and balance their support (Bennett, 1994: 109). A related dynamic is that the industrial parties bargain in the shadow of the law (Bennett, 1994: 111). Strategies when negotiating awards were influenced by the principles that might be applied if a claim proceeded to arbitration, while tribunal members took into account that bargaining was ongoing. The relation between tribunal proceedings and collective bargaining was subtle and complex (Davies and Freedland, 1983: 149–50). These are recurring themes in the literature on the tribunals. For instance, Sykes observed that the award-making process was characterized by a triadic bargaining relationship between employers, unions and the federal tribunals (in Dabscheck, 1981: 439); and the ‘dependency’ inherent in the continuing relationship was a central theme of Dabscheck's work, examined later.

This kind of analysis is a good starting point for the present article. It deepens our understanding of the social foundations of tribunal design and practices, and the relationship between courts and tribunals. Yet, notwithstanding the richness of Bennett's radical critique, it proves unsatisfying as an account of the decision-making process within the main federal tribunals; in particular, of how their structures and procedures mediate the interests of litigants, and how this relates to the tribunals’ functions and the quality of industrial justice. Two points are worth revisiting.

First, Bennett's main focus is on developing an explanation of what interests have been accorded priority, but it is too sceptical of the influence of lawyers and conventional legal values in the tribunals. Take the contention that the High Court thrusts an illegitimate ‘ideology of legalism’ on to the main federal tribunals and the arbitration system (Bennett, 1994: 59). This was allegedly motivated by the historical conservatism and class bias of the judiciary, who were anxious to maintain their own prestige and deny labourist influence in the making of labour law (Bennett, 1994: 74, 77). Here and elsewhere in Bennett's treatment of judicial methods, the rule of law is portrayed crudely as ideology, and the insistence on rule-of-law values is cast as a conservative predilection to exclude the influence of non-lawyers, in particular unionists, from the legal process (Bennett, 1994: 59–61, 105–8). ⁴

This challenge to the liberal conception of the rule of law is misleading. Of course, blind insistence on procedural justice risks drowning out substantive conceptions of justice, but the concept of the rule of law need not be conceived as a cynical device. In Joseph Raz's classic formulation, by contrast, important principles of the rule of law are that laws be ‘prospective, open and clear’, of general application and relatively stable. Courts should be accessible, judges should be independent and natural justice should be observed (Raz, 1979: 214–19). These are internal legal values that serve dignity by ensuring that rules can actually guide human action: an objective that warrants consideration separate from the content of the rules. Adherence is a matter of degree, andinsistence on absolute conformity may be counterproductive where common-sense discretion is suitable (Raz, 1979: 222).

In the Australian context, the importance of formal aspects of the rule of law, such as the independence of tribunal members, adherence to set procedures and the application of precedent, should be explored further. At one point, Bennett (1994: 59) acknowledges that High Court interventions asserting the independence of the main federal tribunals from government may have enhanced the tribunals’ legitimacy and longevity; but she suggests, unconvincingly, that this effect was ironic and unintended by the High Court. Much more is at work. Justice Higgins’ (1922) proclamation of ‘a new province for law and order’ might be taken as referring to the resolution of industrial disputes according to the kind of liberal rule-of-law values mentioned earlier (on the liberal foundations of arbitration, see Macintyre, 1989). The High Court has been determined to uphold such values, even if the practical effect of some judgments was to diminish the prestige of the tribunals, and even if some of the Court's constitutional interpretations are contentious or inconvenient (see Kirby and Creighton, 2004: 126–33; on the Boilermakers case, see also Zines, 2008: 216, 295). ⁵ Further, the manner in which the main federal tribunals have themselves pursued rule-of-law values has garnered respect and imbued the Australian concept of an industrial ‘umpire’ with moral force (for three inspiring papers in this direction, see Kirby, 2004; McCallum, 1996; Munro, 2006).

In the same connection, we should acknowledge Bennett's (1994: 108) concerns about the marginalization of certain groups from the legal process, but reject the contention that court-like procedures necessarily marginalize employees or their unions. Rather, they can guarantee voice. For instance, Sawer (1952: 301) observed that Australian conciliation and arbitration courts ‘with their public hearings, addresses of advocates and cross-examination of witnesses, enable[d] union officials to demonstrate more effectively the dialectics of the capitalist system’. This kind of dynamic deserves further attention.

The second and related cause to revisit Bennett's account is the way it understates the importance of long-standing similarities between the conventional common law mode of trial and the procedures adopted by the main federal tribunals. In Bennett's account of legalisms and the rule of law, ‘judicial’ procedures are portrayed not just as illegitimate in the industrial context, but as ineffective. For instance, she argues that ‘tribunals which take a judicial form cannot perform’ complex roles such the pursuit of industrial and wages policies, because achieving policy outcomes is inconsistent with the judicial ideology of the rule of law (Bennett, 1994: 109).

Again, this is not borne out by the experience of the main federal tribunals. The original institutional investment of the Conciliation and Arbitration Act 1904 (‘CA Act’) was in a ‘court’ of conciliation and arbitration. Since then, the tribunals have deployed recognizably judicial procedures reflecting a delicate, but deliberate, combination of the contextualized and judicial models. The practices of the common law courts have been adapted to the industrial context, in particular to accommodate collective representation and to pursue, with varying degrees of success, industrial policy. Indeed, some highly formalized procedures have proved invaluable precisely because of the complexity of the policy objectives in play. Rather than dismissing approximations of court adjudication in the tribunals as illegitimate and ineffective, parts two and three of this article suggest the value of isolating and tracking those approximations.

The contributions of Dabscheck and Frazer must be dealt with more briefly. For Dabscheck, the regulatory roles of the main federal tribunals are understood in terms of interaction in the ‘Australian industrial relations system’ with other actors (legislatures, courts, specialist tribunals, employers, employees and their representatives). The key to understanding this interaction is the motives of the actors, who are engaged in a struggle for authority (Dabscheck, 1995: 12–17).

In the Australian systems model, tribunals are constrained by other actors in the system, dependent and motivated by a survival instinct. Focusing on the adjudicative role of the CAC and later the AIRC, Dabscheck posed three different arbitral modes, developed from Perlman's (1954a) classic study of the CCCA: first, accommodative/autonomous arbitration; second, administrative arbitration; and third, judicial arbitration (Dabscheck, 1980: 209, 2004: 395; Thornthwaite and Sheldon, 2011). For Dabscheck, the dependency of the CAC and later the AIRC meant that judicial arbitration (an ‘apparent text-book world of reason’ relative to the other more political modes of arbitration; Dabscheck, 1980: 210) could only take place with the consent of the parties involved. He added to his first account of the CAC in a subsequent paper, which attributed more complex dispositions to tribunal members: such as being ‘activist’, seeking to ‘bring their version of justice and reason to the world of industrial relations’ (Dabscheck, 1981: 445).

Dabscheck also identified some important contrasts between the tribunals and ordinary courts: first, the observation that the tribunals are more likely to be involved in an ongoing relationship with those appearing before them, whereas courts are ‘usually involved in an intermittent or one-off relationship’ (1981: 443); second, that ‘courts are more insulated from political influence than are regulatory agencies’ (such as the tribunals) (1981: 443); and third, that the application of the doctrine of precedent is more fluid in the industrial context than in the ordinary courts (1981: 443–5).

While the value of many of Dabscheck's insights cannot be denied, Michelson and Westcott (2001) offer a comprehensive critique of the Australian systems theory. They portray his explanation of tribunal behaviour as ‘crude’ and reactive to contemporary events (2001: 314, sparking an exchange between Dabscheck [2002] and Michelson [2002]). Most importantly for present purposes, I suggest that the Australian systems model does not give us a sure sense of the distinctive features of industrial tribunals.

For instance, it was noted earlier that Dabscheck characterized the CAC and AIRC as ‘dependent’ on the industrial parties. But whether some degree of dependency is surprising given the facilitative roles of the tribunals and their limited resources, or whether dependency is unique to industrial tribunals because of their function and form, is not fully explored. The three contrasts he identified between courts and the tribunals are also debatable, and in any event might be updated to consider the modern functions and modes of the AIRC and FWA.

Considering each contrast in turn, we might say that changes in both tribunal functions and conventions of representation (less peak body representation, more representation by lawyers or self-representation) mean that tribunal adjudication is not necessarily more intimate or ongoing than the interaction that occurs in ordinary courts. In the court system also, government departments, insurers, institutional investors and so on frequently appear before the same list judges and agitate, when dissatisfied with decisions or the cost or process of dispute resolution generally, to change procedural rules as well as the substantive law. As to the contrast between the independence and integrity of courts compared to tribunals, this is potentially very significant; though arguably it was downplayed by Dabscheck, who appears, mistakenly, to have thought that the federal tribunal was constitutionally protected (1981: 443). The independence of tribunal members was, and remains, fragile, a point taken up in the third part of the article.

Finally, in relation to the different approaches of courts and industrial tribunals to the doctrine of precedent, this too is significant. While acknowledging that the role of precedent in civil courts is complex, we might draw out its influence in the tribunals by considering the nature and application of industrial law principles and the touchstone of arbitral decision-making: that matters be decided according to ‘equity, good conscience and the substantial merits of the case’ (now in FW Act s. 578[b]; this is also revisited in the third part of the article). This has implications for the rule of law and may further reveal what is distinctive about the tribunals.

From a different perspective, Frazer also depicts the AIRC as a dependent regulator (2006: 223). He applies legal theories of regulation to the role of the AIRC in supervising the decentralized workplace bargaining framework that emerged from 1992. He notes that the main federal tribunals have generally been conceived as ‘quasi-judicial bodies … influenced by legal norms (both formal rules and informal values) which are not necessarily outcome oriented’ (Frazer, 2006: 223).

When award-making was at its zenith, the role of the commission was direct and creative, setting employment standards. By contrast, the character of AIRC intervention in workplace bargaining after the Workplace Relations Act 1996 (‘WR Act’) was very different. That Act fashioned the AIRC into a ‘meta-regulator’, responsible for a process of self-regulation by the industrial parties (Frazer, 2006: 225). While some powers were exercised by command and control, tribunal members mainly facilitated and supervised enterprise bargaining in a ‘regulatory space’. Direction of the workplace parties and others in the ‘regulatory community’ was subtle and complex and depended on opportunities provided by the legislature for substantive scrutiny, for instance, the ‘no disadvantage test’ at agreement approval stage. Its powers stripped, the AIRC was a ‘weak meta-regulator’ possessing some capacity for adaptation, but generally ‘bureaucratic, legalistic and diffusely structured’ (Frazer, 2006: 238–40).

It may well be valuable to adapt this kind of analysis to FWA, for all its new intricacies. However, while Frazer theorized the existence of a regulatory space and entry points for AIRC influence within that space, in terms of tribunal review powers, I argue that the focus might usefully be shifted onto the distinctive character and quality of adjudicative processes in the tribunals, in terms of procedural design and dynamics. This will also give us an opportunity to instill some meaning in the descriptive compromise `quasi-judicial', adopted by Frazer, above and popular in the literature (e.g. Perlman, 1954b: 213; Naughton, 1993: 3; or `quasi-court' in Cupper, 1976: 351 which was adopted by Dabscheck, 1981: 442).

An Alternative Angle the Principles of Social Ordering

This part of the article highlights the value of a conceptual approach to questions of institutional design. In particular, Lon Fuller's work offers a new way of understanding the distinctiveness of industrial adjudication and the main federal tribunals have developed. Fuller (2001: 61–2) coined the term ‘Eunomics’ to refer to analysis of the principles of social ordering in societies that strive to be ‘just, fair, workable, effective, and respectful of human dignity’. His key article for present purposes is ‘The Forms and Limits of Adjudication’ (1978), published posthumously. One of the most eminent legal philosophers and educators of his day, Fuller took a particular interest in American arbitration, mediation and collective bargaining, having been a negotiator and sometime arbitrator himself (Fuller, 1963; Winston, 2001: 26).

Across a number of papers, Fuller conceptualized different forms of social ordering, or legal processes: legislation, adjudication, mediation, contract and managerial direction (assembled by Winston, 2001). He observed them and idealized their respective core features and competencies, concluding that each process entailed a set of values distinct from the substantive questions at stake. In particular, each process involved a distinctive form of participation.

Bone (1995: 1277) sums up Fuller's conviction and analytical project thus: in complex, dynamic societies, achieving social order requires us to adjust and readjust ‘both substantive ends and legal rules in light of experience with law in practice. And institutions [are] important because they [structure] this ongoing process in an effective way’. This quote gives some sense of Fuller's concern for structure, interaction and feedback in social relations, including relations structured by court or tribunal procedures, as forces in the evolution of law and society. ⁶ (In fact, his project was more fundamental. For Fuller, conformity of decision-making in these processes to the relevant process values had implications for the legitimacy and very legality of the decisions and rules produced. This article will not go so far.)

In a broad sense, there is some correspondence between Fuller's interest in ‘feedback’ and Justice Higgins’ account of the application of reason to industrial problems and the development of early arbitral procedure in the CCCA by experimentation and experience (see Higgins, 1922: ch. 5). I am especially interested in the analysis of adjudication to explore how the procedures of tribunals embody a response to the functions allocated to them. Fuller sought to define adjudication not by its output, but by its characteristic procedures and values contrasted with other processes, in particular contract and voting. Adjudication affords parties to a dispute the opportunity of participation according to defined procedures and subject to a requirement of rationality. Submissions are judged by adjudicators who are impartial, independent and trained in legal methods. The process is ‘rational’ in that the parties must refract their positions through principles already enunciated in statute and case law; a rationality different and more urgent for adjudicative processes than for decision-making in the other forms of social ordering (Fuller, 1978: 367).

Further idealized features of adjudication follow from this distinctive participation value. Adversarial advocacy is necessary to direct the views of the parties into rational discourse: ‘The arguments of counsel hold the case … in suspension between two opposing interpretations’, also guarding against a tendency that might afflict the adjudicator ‘to judge too swiftly in terms of the familiar that which is not yet fully known’ (1978: 383). The adjudicator should also give reasons, so that decision-making is transparent and can be reviewed. Fuller's emphasis on lawyers as ‘guardians of due process’ (1978: 384) and his critique of inquisitorial procedures (the arbitrator acting on her own motion) are not recited here, except to say that it may be a fair charge that he underplayed the role of the arbitrator. Allison, for instance, has suggested we might focus more attention on the ‘purposive interaction between the adjudicator and the parties’ (2000: 203, 205).

Fuller also attempted to explain the effect of an adjudicative process on its subject matter. It is worth setting out in full because of its interest to that long-standing question in Australian industrial law: whether arbitrators were/are engaged in a judicial function (deciding claims of right) or a legislative function (resolving conflicts of interest):

 Adjudication is a process of decision that grants to the affected party a form of participation that consists in the opportunity to present proofs and reasoned arguments;

It is through this process that the tribunal becomes involved in ‘articulating the implications of shared purposes’ (1978: 378, 381). Adjudication thus givesexpression to two broad principles of social order: reciprocity and organization by common aims (1978: 363), in that parties with divergent interests are bound together by rules and principles built on common ground (1978: 381). ⁷

The next question for Fuller was the kinds of tasks and functions to which adjudication is best suited. He cautioned against the allocation of responsibility for resolving ‘polycentric’ disputes to adjudication. Polycentricity refers to problems that affect interests beyond those of the disputants. To explain the concept, Fuller made analogy to the unpredictably variable distribution of tension through a spider's web when different strands were pulled at, or to allocating positions on a football team. For instance, in the industrial context, decisions in wage cases clearly pose polycentric problems because they have systemic labour market and social effects. He contrasted this to the normal binary adjudicative tasks he thought were characteristic of common law courts, for example, interpretation of the terms of private law contracts (1978: 395). ⁸ Deploying adjudicative modes to resolve polycentric problems is risky because, according to Allison (2000: 193), ‘As a result of inadequate participation, the adjudicator is inadequately informed and cannot determine the complex repercussions of a proposed solution’.

Fuller offered a ‘sketchy and suggestive’ analysis of ‘mixed, parasitic and perverted forms of adjudication’ (1978: 405–6). Certain American examples of tripartite labour arbitration, for instance, mixed contract and adjudication (1978: 395–7). Perhaps surprisingly, his use of the terms ‘parasitic’ and ‘perverted’ were not intended to be pejorative. As Bone (1995: 1300) notes, the analysis was cautionary rather than prohibitive, encouraging careful and informed design of institutions, balance between institutions, and justification for departures from certain process and rule-of-law values (see Barber, 2001: 80; Fuller, 1978: 405–6; but contrast with Allison, 2000: 194, 198; Fuller, 2001: 62).

Fuller's cautions were addressed to constitution-makers and legislators, who might create an adjudicative forum, allocate it functions and set out its rules of procedure. They were also addressed to tribunal members, who may have scope to determine the composition of a bench, its methods and procedures. Indeed, reflecting on his time as the second President of the CCCA, Justice Higgins remarked: ‘I had to learn the business with no book of instructions, no teacher other than experience, no kindly light except for the pole star of justice’ (1922: v). This suggests a scope for discretionary judgement as to standards and procedures, which would be unfamiliar to modern tribunal members. Stewart (2004: 255), for instance, noted the ‘dramatic increase’ in the complexity of regulation of the work of the AIRC and the ‘fall in the degree to which it [was] trusted to do its job’.

Function, Form and the Evolution of Industrial Adjudication

In this section, Fuller's concept of adjudication is contrasted with the structures and procedures of the main federal tribunals in order to draw out what is distinctive about industrial adjudication. ⁹ The particular adjudicative functions of the tribunals under consideration include award-making upon an industrial dispute, wage-setting, test cases to vary award standards and, of modern importance, certification of voluntary agreements, workplace bargaining applications, protected industrial action ballot orders and unfair dismissal complaints. This is admittedly a diverse group. There is no single mode of adjudication in the tribunals and different functions call for different methods.

We might first question whether it even makes sense to conceive of these functions as adjudicative, given that, perhaps with the exception of unfair dismissal cases, they involve resolving conflicts of interest rather than rights claims in the conventional juridical sense. Some accounts have looked beyond the trappings of tribunal procedure to contend that the historically dominant functions of award-making and wage-setting saw the tribunals act more like economic legislatures, laying down minimum standards, than adjudicative bodies (Hawke 1956 on the basic wage; see also Sawer 1952: 302, 310; Perlman 1954a: 32, ch 2; Perlman 1954b: 208. Similarly Kahn-Freund, in Davies and Freedland 1983: 150, viewed awards as a kind of `legislative technique').

These accounts have explanatory value in relation to the systemic considerations and effect of tribunal decisions, but they do not deny the presence of adjudicative procedures. The deliberate institutional investment in tribunalism, followed by the design of procedures and the accumulation of industrial law principles, was to convert conflicts of interest into claims of right, to be resolved rationally. Adapting Fuller (1978: 369), ‘This conversion [was] effected by theinstitutional framework within which both the litigant and the adjudicator function[ed]’. In this sense, even the old national wage case and award-making processes seem to have been more like those in courts than a legislative process (see also McGarvie, 1965: 90). ¹⁰

To complicate matters, however, it may be that the ‘economic legislature’ tag is more appropriate now than ever, to describe the procedures of FWA's Minimum Wage Panel (FWA-MWP), which is a descendant of the Australian Fair Pay Commission rather than of the AIRC. The wage adjustment process is no longer triggered by an application. FWA-MWP invites submissions, holds consultations and commissions research and modelling to aid its deliberations (FW Act s. 590[2][g]). Similarly, during the award modernization process, FWA was directed by ‘requests’ from the minister, then published drafts for comment and issued final awards (see Creighton and Stewart, 2010: 6:17; Stewart 2011). These modes can be seen as a dramatic reformulation of the regulatory response to problems of polycentricity, and see FWA members acting more like a legislative committee or like law reform commissioners than tribunal benches past. They are indeed legislative techniques and are markedly different from the adjudicative modes retained for functions like unfair dismissal cases and workplace bargaining applications. FWA wears so many hats now, that Stewart (2011) describes it as an agency rather than a tribunal. Again, however, this article is concerned with its arbitral functions and it is safe to refer to FWA as a tribunal for this purpose, which may come as a relief to many of its members.

Four clusters of structural and procedural features are now discussed, to contrast the mixed adjudicative processes of the main federal tribunals with Fuller's idealized features of adjudication: how the parties are engaged; tribunal composition and independence; the use made of previous decisions; and procedural flexibility.

These are important categories for isolating and re-examining the evolution of the tribunals, drawing on the existing rich literature and the discussions in the first and second parts of the article, and suggesting new lines of inquiry.

Conclusion

Justice Kirby has observed that the ‘process of creation and renewal [of the industrial tribunals] has, like a human life, been one of constant rediscovery of things that are fundamental’ (2004: 229). Industrial arbitration was intended as a process of reasoned deliberation between employer and employee interests, and the continuing investment in an `umpire' reflects a determination to resolve conflicts of interest according to principle. The CCCA was originally styled as a court and its successor tribunals, including now FWA, have in significant ways approximated court adjudication. But how, and what has been/is important about those approximations? It is worth exploring, with the aid of Fuller's work, how arbitration in the federal industrial tribunals has evolved over time to accommodate functions allocated by the legislature and tensions associated with those functions. Towards this end, certain departures from, and innovations on, Fuller's adjudicative ideal were identified in the third part of the article.

We might explain those departures and innovations in a number of ways. Substantively, the conviction informing the continuing use of industrial tribunals today is that power imbalances in employment relations can and should be regulated in the public interest. This makes an institutional and a procedural difference (adapting Allison [2000: ch. 3] on English public law). This may help explain tribunal features such as the support for peak body representation and the disposition of the arbitrators (perhaps more interventionist and inquisitorial than in Fuller's adjudicative ideal). Further, as the institutional investment in tribunal arbitration has evolved, an important consideration has been how to tackle inherent problems of polycentricity, which was highlighted a number of times in the third part of the article.

There is already a rich literature on the main federal tribunals, yet more research remains to be done. It might be that their procedural innovations are considered not just sound, but ingenious, and that Fuller's ‘ideals' are unduly limiting. But it might also be that we discover worrying patterns in the AIRC's and FWA's accommodation of successive parliaments’ instructions to adopt more and more informal procedures. This raises a number of challenges: to explain and justify tribunal design and redesign by reference to procedural values and best traditions; and to understand better the nature of the modern tribunal's contribution to the rule of law in employment relations.