Tribunals in the System of Government

Peter Bayne

Tribunals in the System of Government (PDF 101KB)

Tribunals occupy a significant, distinct and an independent position in our systems of justice and of administration. One example at the Commonwealth level is the Administrative Appeals Tribunal (AAT) and another the Social Security Appeals Tribunal (SSAT). SSAT decisions affect the way in which the Social Security Act applies to the very many Australians who depend on pensions and benefits provided for under that Act. The AAT can hear appeals from the SSAT.

These kinds of tribunals, at first glance, look and act like courts. They receive and evaluate evidence - often through an oral hearing. There will be parties to the proceedings of the AAT; in respect of an unemployment benefit claim for example, there is on the one hand the Government agency - the Department of Social Security and on the other there is the claimant. The physical surroundings in which the Tribunal operates will look very much like a court room if we are looking at the AAT somewhat less so, indeed, a great deal less so, if we are looking at the SSAT. But even in relation to the SSAT, in most cases the parties will gain an impression that the Tribunal will listen to the evidence and find facts which, of course, is what courts do. The parties may also perceive that the Tribunal will apply the law to those facts; that is, the law in the Social Security Act.

Let us notice a couple of things about that law. First, the terms of the relevant statute may be very vague so that a choice is left to the Tribunal to say, in effect, what the law means. For example, a decision about whether a person is qualified for an unemployment benefit may require, depending on the case, an answer to the question whether the claimant had taken reasonable steps to obtain work. That is indeed one of the key questions concerned in the application of the unemployment benefits provisions of the Social Security Act.

The second thing we should notice about that kind of case is that the law applied by the Tribunal will, in many cases, have been applied already by the government agency concerned. Officers of the Department of Social Security will have made a judgment about the facts and the law of a particular case. They will do that in a somewhat distinctive way. They will not look so much at the law, at the terms of the Act, but at a statement about what the relevant provision means which will have been prepared by other officials who work within the agency. So that in the unemployment benefit case the officer will have looked at the unemployment benefit manual to see what it says about what taking 'reasonable steps to obtain work' means.

An important point to make about this process is that the manual statement will often be a part of a much larger picture of what the Government has in mind about how the law will work. So the policy¹ about the unemployment benefit will take into account, for example, the size of the Government's outlay for that benefit and for social security in general. It may also take into account just what the Government hopes to achieve by making the unemployment benefit available, say in the way of restructuring the labour market. So the manuals - which on the face of it simply give an interpretation of the provisions of an Act - are framed having regard to not only the terms of the Act, but also to policy on a much wider range of issues which might not be apparent from looking at the terms of that Act.

Before I examine the role of external review tribunals (such as the AAT and the SSAT) when they hear appeals from decisions, it is necessary to look briefly at the role of the courts when they review administrative action. We have to bear in mind throughout this discussion the contrast between the different roles of a court and a tribunal. The courts have for centuries decided cases concerning challenges by citizens to government. Lawyers, when they talk about court review, do not normally talk about an appeal. Rather, they say that the court reviews a decision. This distinction is elusive and although it is one that is central to the argument I will simply have to state it rather baldly at this point.

A review considers only whether what the decision-maker did was lawful in the sense of being within the scope of her or his lawful authority according to the facts as found by the decision-maker. The court does not ask whether it would have come to the same decision as the decision-maker. This allows that one decision-maker may differ from another, yet both may be acting lawfully. In contrast, a tribunal such as the AAT, when hearing an appeal, will consider not only this question of legality but also whether what the decision-maker did was the correct decision to be made. Indeed, one can go further and say that in many cases, the Tribunal will not be much concerned about what the decision-maker did; it will simply make its own decision about how the relevant law should be applied to the facts as it finds them to be. It is important to see that tribunals act in this way, because it reveals that their decisions will have a much more direct influence on the course of administration by decision- makers; a much more direct influence, that is, than have the courts.

The influence of a body like the AAT is also more pervasive because it is usually much easier and cheaper for aggrieved citizens to appeal to a tribunal than it is for them to seek review by a court. It is for that reason that a tribunal is more frequently called upon to hear an appeal than a court is called upon to hear a review.

With these matters in mind let me proceed to state the overarching perspective of the paper. It is this: if we are to gain a proper appreciation of the role of tribunals and their place in the system of government we need to see that they occupy a place in both the system of justice and the system of administration.

We can view tribunals as part of the system of administration because they decide appeals in the way I have just described. The statutes which they administer require the making of judgments which call for choices which must reflect a conception of the social and economic objectives which the statutes support. The cumulative effect of their decisions determines how questions of public policy are answered; for example, just who gets the unemployment benefit, which in turn will affect matters such as the restructuring of the labour market. Thus, we see them as part of the system of administration because an appeal tribunal is doing exactly what the primary decision-maker does. We can also, however, see them as part of a system of adjudication because in deciding particular disputes about these matters they act, and are seen to act, like courts; that is, by the impartial application of the law to facts ascertained through a procedure which is essentially adversarial.

It is my argument that if we take the standpoint that tribunals have a dual role, we might get a better answer to a range of questions concerning them - for example: how they should be composed; what the tenure of their members should be; how they should relate to the courts; and how far they should take account of government policy and administrative understandings of the law in the interpretation of the statutes that they administer.

I cannot address all of those matters here. Rather, I will consider the quite fundamental matter of how a body like the AAT should interpret statutes. The answer to that question will then be related to the notion of the independence of tribunals.

Critical to the argument is the notion that a tribunal has a normative role in relation to the agencies in respect of whose decisions it hears appeals. A decision of a tribunal (or a court) has a normative role to the extent that it has an effect on agency behaviour which travels beyond the case at hand and shapes the way an agency deals in the future, not only with cases of that kind, but with other cases where, by analogy, it would be appropriate to apply the tribunal's statement of principle to justify its decision in an earlier case. The point was made by Brennan J when, defending the lawyer's contribution to review of administrative action through the processes of the AAT, his Honour said that review is 'normative, improving primary administration by defining the nature and extent of the administrator's function ...'.²

Let me try to explain it this way. When a court reviews the legality of a decision its focus is on the narrow question of the legality of some administrative action. It can be expected that the agency will in future decision-making respect the court ruling, but that ruling is only partial in its effect on the agency because it does not touch upon what kind of decision the agency should make that is within the legal boundaries of the statutory power to take the action concerned. The court's ruling is partial because it goes only to the legality question. Furthermore, having regard to the whole range of administrative action taken by the agency under the statute, and the infrequency of judicial review, court rulings are sporadic.

Now, in contrast, the AAT's decisions affect not only the agencies' understanding of the legal boundaries of their statutory powers, but deal also with what kinds of decisions they should make within those legal boundaries - that is, with the kinds of decisions agencies should make, taking into account all matters and not questions of legality. Thus the decisions of the AAT could have a considerable normative affect on an agency because they establish what is correct action by the agency rather than only what is lawful. Tribunal decisions are also more pervasive than sporadic because they range across a wide spectrum of agency activity. Tribunal decisions are also likely to be more frequent than court decisions, for the very important reason that it is much cheaper to appeal to a tribunal.

To come now to the question of how a tribunal like the AAT might interpret a statute. The business of the AAT is interpreting statutes which confer administrative power. This might be thought to be a rather straightforward matter - that it is simply a question of reading the words of the statute in their context and in the light of the purpose of the Act concerned, and producing an answer. I want to suggest that it is much more complex than that.

When Parliament enacts a statute, such as the Social Security Act, and confers powers on an agency's officials to determine questions such as whether someone is taking reasonable steps to obtain work, what it is, in effect, doing is conferring a role in the law making process on that agency. This is most obvious where the statute confers what we can call an open-ended discretion. For example, there is a provision in the Social Security Act which says that the Secretary may waive the recovery of an overpayment. That is all it says. There is no indication of the circumstances in which waiver might be appropriate or not. When the Secretary formulates a policy, in the sense of a guide to action to govern when the department will and when it will not waive recovery of overpayments, the Secretary is, in effect, rewriting that section of the statute by giving it a content which, just looking at the words of the statute, it does not have.

That process also occurs, it seems to me, where the statute is more precise but nevertheless requires the making of a qualitative judgment on a matter about which reasonable minds could differ; for example, on the question of whether a person has taken reasonable steps to obtain work. My point is this: where there is some vagueness in the language of the statutory provision the interpreter must choose from a range of possible meanings, the one he or she considers are conveyed by the words of the provision. Words such as 'reasonable steps to obtain work' do not compel as a matter of deductive logic the meaning which the interpreter gives to them.

Let me give you an example of this, because it is central to the argument. Take the question whether people have been taking reasonable steps to obtain work if, rather than looking for a job, they have sought to train or to educate themselves for some form of a job - such as a computer programmer - for which they are not presently qualified. It can be quite plausibly argued both that such a person has and has not taken reasonable steps to obtain work. The choice cannot be made by simply contemplating the words of the statute. Rather, it has to be made by reference to other considerations; (the 'inarticulate major premise' as it has been put).³

An obvious consideration is the purpose of the unemployment benefit. Is it legitimate to see it as a support for training and education for employment? Reasonable minds can differ about the answer to that question. The claimant could say: 'I am looking for work if I am trying to retrain myself to get a job for which I am not presently qualified, in particular if it is a job in a sector of the labour market in which there is a shortage of labour. I might be able to get a job digging up the roads, but there is plenty of competition for that kind of job, and I have a better chance of securing employment if I train myself for work as a computer programmer where there is a shortage in the labour market. I can moreover make a better contribution to national economic development as a computer programmer'. The thrust of this kind of argument is to say that a person is taking reasonable steps to obtain work if they are seeking to retrain themselves. The Department could, of course, counter with the argument: 'Well, you could possibly find a job as a road digger if you went out and looked for it. That is what taking reasonable steps means - trying to find a job that you could possibly get. It does not encompass retraining yourself'. This kind of debate is rarely solved by looking only at the words of the statute.

Thus, my argument is that for the purpose of assessing the relevance of some policy of an administrator or an official to the interpretation of an administrative power it is legitimate to assimilate the classic administrative discretion (such as whether an overpayment should be waived) to the kind of judgment that must be made in the application of a vaguely worded provision which defines rights and duties (such as whether a person is entitled to unemployment benefit because they have taken reasonable steps to find work). In both sorts of case, the agency has a law-creating role.

In performing this role, the agency should not approach the matter in the same way as the Parliament does. It simply cannot say that what Parliament has said is undesirable and should be replaced by some other verbal formulation. The official must accept and work from the statutory text and must give meaning to that text in the framework of the whole statute (and perhaps of other statutes).

But that framework very often does not yield a clear answer, and the agency will often seek its inspiration in whatever government policy exists in relation to the purpose of the statutory provision in question. That is what administrators should do and it is what they usually do. They see their job to be to apply those sorts of provisions in the context of government policy, whether it is policy in relation to that provision or, perhaps, in relation to a broader policy such as the budget allocated to the program to which the statutory provision in question relates.

Generally (although with few exceptions) the courts will not look at a statute in this way. There is some justification in this approach because they are concerned only with the question of the legality of administrative action and they are clearly not part of the chain of administrative decision-making. (Whether in the end this approach is justified is beyond the scope of this paper).⁴ The courts are very chary about giving recognition to whatever government policy may exist in relation to the interpretation of a statute. But appellate tribunals, which remake the decision under appeal, are part of that chain of administrative decision making. How they approach the interpretation of statutes gives an important part of the answer to the question of where they stand in the system of government.

There is another matter to mention here. Some of you are probably aware that the Acts Interpretation Act 1901 (Cth) has been amended to make it clear that the courts and bodies like AAT may look at legislative history when interpreting a statute. But this history is not often a particularly helpful guide to sorting out questions such as what it means to take reasonable steps to obtain work. That is partly to do with the Tribunal's attitude towards legislative history, but largely to do with the fact that the history is often extremely unhelpful on just those sorts of precise questions.

The AAT could turn to the sorts of documents which the administrators use - that is, it could look at an agency's understanding of the meaning of the statute which it administers, and this brings us to the critical issue to be confronted. That sort of understanding is found in documents like manuals, which are there for the guidance of officers of the agency but which are also available to the public under the Freedom of Information Act 1982 (Cth). Also available might be a ministerial statement of policy. It might be only of the goals of a program or it might be more specific and indicate an understanding of how a discretion should be exercised. It might have been circulated only within the agency, or it might be tabled and debated in Parliament. But, by and large, if you are looking for an administrator's understanding of policy, it will be found in a manual or in some other document which probably has not seen the light of day in any parliamentary proceeding and which may not have passed over a minister's desk.

The question is: what weight should the AAT give to those sorts of administrative understandings of the law? In dealing with that question, the Federal court and the AAT have drawn some distinctions. In the first place, they give the greatest respect to ministerial policy statements, particularly when the discretion is open-ended and is directed to a topic upon which the AAT feels it is not suited to adjudicate, such as the question of when people convicted of certain serious criminal offences should be deported. It is in relation to that sort of case that Justice Brennan said that the AAT should 'favour only cautious and sparing departures from Ministerial policy'.⁵ But faithful to the decision of the Full Court of the Federal Court in Drake v Minister for Immigration and Ethnic Affairs,⁶ Brennan J spelt out the limits of this approach. It was said that

a policy must be consistent with the statute. It must allow the Minister to take into account the relevant circumstances, it must not require him to take into account irrelevant circumstances, and it must not serve a purpose foreign to the purpose for which the discretionary power was created. A policy which contravenes these criteria would be inconsistent with the statute ... Also it would be inconsistent ... if the (Minister's) policy sought to preclude consideration of relevant arguments running counter to an adopted policy which might be reasonably advanced in particular cases. The discretions reposed in the Minister by these sections cannot be exercised by binding rules .... His discretion cannot be so truncated by a policy as to preclude consideration of the merits of specified classes of cases.⁷

This statement has been frequently cited, and applies to any kind of exercise of an administrative power by reference to a policy.

But where the discretion is vested in an administrative body or official, and, in particular, where what is involved is not a discretion but a discretionary judgment, the AAT has not been willing to accord weight to the decision-maker's understanding of the statutory provision. It is in that sort of case that the AAT is likely to say that government policy must yield to the words of the Act.

This approach suggests that the AAT can find an answer in the Act, but in very many cases, this is not so. Contemplation of words like 'reasonable steps to obtain work' will not yield a clear answer to the question of whether a person seeking to retrain herself or himself falls within the category of having taken 'reasonable steps'. The question of how far administrative policy should figure or should be given weight in the AAT's interpretation of statutes is one that needs a somewhat more sophisticated answer than the one the AAT has given.

Just how far the AAT might go about giving weight to administrative understandings is a matter which is quite complicated and requires a sophisticated answer. Only a few elements of the analysis can be touched on here. In the first place, it is necessary to deal with an argument made by the Tribunal that if it were to approach a question in a way different to the Federal Court it would be doing a disservice to the parties because its decision would be simply overturned by the Federal Court on appeal.⁸ There is obviously force in this argument, and a full reconciliation of administrative understandings of the law with the AAT's approach probably requires a change in attitude by the courts to this same matter. But the AAT may not be as hamstrung as may appear at first glance. (Of course, many people who resort to the tribunal - other than the government, the wealthy and those on legal aid - could not afford to appeal to the Federal Court. The AAT may also be concerned about how it would be seen by the judges.)

The court is concerned with a question of law, which in a case where the legal limit of a discretion or a discretionary judgment is concerned, can be sufficiently described as a question about whether the limits to the application of ministerial policy spelt out in Re Drake (No 2)⁹ have been observed. Within these limits,¹⁰ the court should not be concerned that the AAT has given weight to the reasons why the administrative decision-maker took a view that certain considerations were relevant or not relevant to the exercise of a discretion or the making of a discretionary judgment. I will attempt to explain this distinction.

There are two situations in which the AAT will need to take into account a Federal Court response. The first is where the court has given a ruling on the point in issue. The second is where by the orthodox approaches to statutory interpretation it is plain that the administrative agency is in error. But this leaves aside a great many cases, such as the question of whether a person who is retraining is taking reasonable steps to find work.

To explore what the AAT might do by way of taking into account the administrator's understanding of the law in this sort of case, we need to return to the distinction between judicial review (the function of a court, and which also describes its function when hearing an appeal on a question of law from the AAT), and an appeal on the merits (the function of the AAT). One¹¹ important way that distinction applies is as follows. Where a statute confers a discretion, a court exercising judicial review will consider whether the administrative agency had regard to the relevant considerations and left out of account irrelevant considerations. It is not concerned with whether, within that framework, the agency balanced the relevant considerations in a manner with which the court agreed, or, to put it more simply, the court will not intervene even if it thinks the outcome is wrong. (This is subject to some qualifications which need not be considered.) There are two levels of choice in this sort of case: (1) what are the relevant considerations and the irrelevant considerations, and (2) within this framework, what is the best decision?

Now when this sort of case comes to the AAT on an appeal from a decision-maker, the Tribunal does not need to be concerned that the Federal Court may come to a different view at level (2), for that is not a concern of the court. At this level, the AAT can have regard and give some weight to the decision-maker's reasoning and any policy that may have played a part in that reasoning.

It is at level (1) that questions arise which are the concern of a court hearing an appeal from the AAT. Where the question concerns which considerations are irrelevant, the touchstone is to consider the subject matter, scope and purpose of the statute 'in order to determine if there is to be found some implied limitation on the factors to which the decision-maker may legitimately have regard'.¹² Similarly, '[i]f relevant factors ... are not expressly stated, they must be determined by implication from the subject matter, scope and purpose of the Act'.¹³ These tests leave a great deal of choice to the AAT (and to the court), and what the AAT needs to be concerned with is whether its choice is within the range of choices that the court might make; it cannot do any better than this. What it probably cannot do is to say that it simply defers to the decision-maker's assessment of what is relevant and irrelevant, but in theory I cannot see why it could not give some weight to this assessment in the same manner that Brennan J said that the AAT should favour only cautious and sparing departures from ministerial policy.¹⁴

The question now arises: why should the AAT take that view? I suggest that it should for a number of reasons which come back to the conception of the AAT as a part of the administrative system of government as well as part of the system of adjudication.

The first point to make is that unless the AAT is prepared to approach the exercise of discretions and the making of discretionary judgments in somewhat the same framework or frame of mind as an agency, then its decisions are less likely to have that normative effect on agency decisions which are rightly said to be an important dimension of its work. I say that for a couple of reasons. First, while it is usually the case that the agency will attempt to mould its future conduct to conform to an AAT decision, there is always the risk that the agency will simply ignore what the AAT says. If it does that, the AAT's normative role is frustrated. (This has clearly happened in a number of areas of social security and customs administration, and generally across several agencies so far as the application of the Freedom of Information Act 1982 is concerned).

Beyond defiance there is another step. It has become apparent in recent times (and this is particularly true of the Social Security Act) that the Government is simply changing the law to overcome AAT decisions. Straightforward amendment of the Act to make a new rule to apply in the situation dealt with by the AAT is one option and on the face of it is hardly objectionable (although where amendment of this kind becomes very frequent the AAT's normative role is frustrated). Of greater significance is the now frequently used device of inserting a provision in the Act concerned which enables the Minister or the departmental head to make a kind of delegated legislation (variously called 'principles', 'directions', 'guidelines', and the like) whereby the scope of a discretion or the parameters of a discretionary judgment can be refined by that delegated legislation. In this way, the Minister or the departmental head can not only reverse the AAT decision(s) which prompted amendment but can, from time to time without the need for further recourse to Parliament, reverse future AAT decisions. (The exponential growth in this form of delegated legislation has ramifications well beyond the point I make here).

The point is that these kinds of amendments remove discretions and replace them with rules and when that occurs the normative role of the AAT is frustrated. Whereas in the past it would have been able to exercise a discretion which took account of the complexity of circumstances that surround the sort of case being dealt with, it must often-times now say (or perhaps think) 'I am sorry. You may feel hardly done by and we sympathise, but our hands are tied by the precise terms of the law. We cannot do anything for you'. That is often the result when discretions or discretionary judgments are replaced with very precise rules. It seems to me that this is happening partly (indeed largely in some cases) as a response to the fact that the AAT has adopted an interpretation of the statute which pays no attention to the administrator's point of view.

There are a number of other reasons why the AAT should give some weight to an administrator's understanding of the law. In the Re Drake and Minister for Immigration and Ethnic Affairs, Justice Brennan argued that some weight would be given to ministerial policy. He said that that was appropriate because for the Tribunal to do otherwise might nullify 'any mechanism of surveillance which the relevant statute permits or provides. To depart from ministerial policy thus denies to parliament its ability to supervise the content of the policy guiding the discretion which the parliament created'.¹⁵ His Honour said that in the context of a ministerial policy which was tabled in Parliament and his point was that, that having occurred, one could say that for the Tribunal then to depart from that policy 'denies to parliament its ability to supervise the content of the policy guiding the discretion which parliament created'.¹⁶ His Honour attached a great deal of significance to the fact that the ministerial policy had been tabled in Parliament.

I suggest that that argument is not limited to policies made in the name of Ministers which have been tabled in Parliament. Agency policies are often designed to carry out an objective set by the Government and it is of course difficult to distinguish ministerial or cabinet policy from administrative policy. Whether or not that is the case, Parliament, through the questioning of the Minister, can call the Minister to account for departmental policies (although there are obvious limits to its ability to do so and this line of argument is not very strong). What I suggest then is that if one is to accept Brennan J's argument then there is also some point to giving weight to administrative policies because they do not stand in a substantially different position to the sort of policies that are announced by a Minister and tabled in Parliament.

Apart from that reason, there are a number of functional reasons why the AAT might give some weight to agency interpretations of the law. An agency has a relatively greater capacity to develop policies for the exercise of administrative power which will enhance the purpose of the scheme to which the statutory provision relates. For example, agency policy about the unemployment benefit will be part of a larger policy which looks at the role of other benefits under the Act, such as the invalid pension. It will also have regard to the Government's objectives, say, in relation to the restructuring of the labour market, or the size of the Parliamentary appropriation. The agency is better suited than the AAT to see the bigger picture of how the interpretation of a particular provision of an Act will affect those larger questions.

One can also say that agency policy is more likely to enhance consistency in the application of the statutory provision. There are well known examples of AAT panels differing between themselves about, for example, when an invalid pension should be granted. The agency is in a position to ensure greater consistency of application. It is also in a position to ensure a geographical uniformity in the application of the particular provision so that one does not get a different view about who gets the invalid pension in Perth to who gets one in Melbourne. Yet, with a tribunal making up its own mind on those questions and paying no attention to the agency view, the risks of inconsistency are greater.

To recapitulate, what I am suggesting is that because the AAT has a role in the system of administration and because what it does is very different to what a court does, it might reconsider its attitude to the question of what weight should be given to administrative interpretations of the law in the interpretation of statutes that underpin administrative decision-making.

This brings me to another matter which is germane to my argument. Some of the considerations which I have relied on to support the argument point to the significance of Parliament as the law creating body and the body which calls agencies to account. It was certainly a matter that Justice Brennan thought important in Re Drake. This sort of argument - that by giving weight to agency policy you will be recognising the role of Parliament - has much greater force where it can be seen that during the legislative process there was a statement of the policy which would guide the agency in the exercise of a particular administrative power.

Until recently there was not much of an opportunity for an agency policy of this kind to be stated in the course of parliamentary proceedings. Under the normal procedure, when a Bill is debated by the Committee of the Whole, it is debated clause by clause and the only person who speaks for the government is the Minister (or, in the other chamber, a representative of the Minister). He or she may not have a great deal of knowledge about the particular purpose of a provision that confers a discretion (assuming of course that a Member raises any query about the matter). Agency officers sit near the Minister or the representative to advise and to relay information that will enable a reply (which may not be an answer!) to be given to questions. This process is unlikely to yield an appreciation of how the government expects a particular provision to work (such as how a discretion will be exercised). Apart from the Minister's lack of expertise, there is simply not enough time to address that kind of problem.

There has, however, been a major change in the way the Senate considers bills which could provide an opportunity for an agency to articulate its policy about how a particular statute is meant to operate. It will also give the Senators some opportunity to impress (and the government to explain) their perception of what policy should guide the application of the statute. The Senate has passed resolutions to introduce a procedure whereby, following the second reading speech of a bill, it might be referred to the appropriate Legislative and General Purpose Standing Committee of the Senate for report in detail on its clauses. What is envisaged is that, after the second reading, a bill will be referred to one of the subject committees which will hold hearings on the referral of bills. That may or may not replace consideration in a Committee of the Whole of the Senate but it is hoped in time, in some cases, that that will be the case.

The report of the Senate Select Committee on Legislation Procedures, which recommended this procedure in December 1988, suggested that a subject committee might consider whether the policy of the Act is coherent, whether its provisions are effective and, indeed, the most effective means of carrying out that policy. In the hearings it will hold, the subject committee may call for public submissions and it may hold hearings by taking oral evidence. The Senate select committee suggested that Ministers make available to the subject committees agency officers who can provide advice on the technical aspects and background of the bill. The guidelines on conduct of public servants allow them to say what government policy is, but they are prohibited from announcing their own view on policy. One can imagine that if a Senator picks up a provision in the Act which confers a discretionary judgment on an official he or she can say, 'What does that mean? What do you have in mind? What does the agency think "taking reasonable steps to obtain work" means?' The agency answer will then become a matter of public record and parliamentary record.

In a number of ways this process might have significant ramifications for a number of reasons. One of them is germane to the topic that I am considering. It seems to me that, to the extent that that kind of record emerges from this new procedure, the AAT will have a firmer basis for linking agency policy to parliamentary supervision and control. It may provide a further reason to give some weight to an agency interpretation of the law.

Let me now come back to the independence of the AAT. It might be said that the AAT ceases to be independent if it gives weight to agency understandings of the law. In considering this question of independence of tribunals, Justice Brennan has argued that one should see them not as part of the administrative machinery of government but as part of the adjudicative system. He said that if the review process is intended to be normative the lawyer's contribution is 'dispensable and salutary'.¹⁷

One could have no quarrel with the notion that the AAT is independent but, with respect, I think that Justice Brennan may have overstated the matter. The AAT, it seems to me, is independent in the way that a public servant is not. It has been said that the administrator of public services 'is an employed person par excellence ... that employment involves a subordination to higher authority, a liability to receive instructions as to the work to be done and the manner in which it is to be performed - a liability which pervades the entire realm of public administration'.¹⁸ The AAT is not dependent in that sense - it is independent in the sense that it is not given instructions and the like. It does not compromise its independence by interpreting statutes in the manner which is both more characteristic of the administrator's approach and gives some weight to a particular illustration of that approach which is relevant to the matter before the Tribunal. It may narrow somewhat the circumstances in which it will come to a decision different to that of the primary decision-maker, but its independence is not compromised.

The perspective that the AAT is as much a part of the administrative as the adjudicative system could well be applied to other aspects of tribunals - to questions of tenure, of procedure, and of their jurisdiction. But at this point I will cease.

Discussion Period

Question:

It seems to me that the tenor of your comments suggests two concerns: one a concern about the increasing drift towards a judicial or quasi-judicial status in the Administrative Appeals Tribunal which is unwarranted by the concept that lies behind it, and secondly a concern about the ministerial responsibility implications, the implications for the principle of ministerial responsibility in the kind of independence which the Administrative Appeals Tribunal has harnessed to itself. Would you like to comment on that?

Response by Mr Bayne:

I think it is a fair statement that those concerns are reflected in the paper. I suspect that the problem has been with the AAT, that it has seen itself as largely part of the judicial or adjudicative system of Government. It has done that, I think, in order to assert its independence from Government, but I am not sure it needed to go to the extent of assimilating its activities to that of a court. The point is commonly made by looking at the procedure the AAT follows, and one could say a good deal about that. Certainly, if one practises before that Tribunal, it is very difficult to see an enormous difference between the way a court operates and the way the Tribunal operates. I am seeking to say that the consequence of the AAT seeing itself as part of the adjudicative system - and one can understand it did that to distinguish itself from the administration - is that it overlooks the fact that it plays a very different role in relation to the administration than does a court. The AAT has a normative role, as I explained. If it is going to have that normative role, it has to acknowledge the administrator's viewpoint on the interpretation of the law to a greater extent that it has done.

There is a common criticism that the AAT is too legalistic, too much like a court. In defence of the AAT, that role is often thrust upon it by the nature of the legislation with which it is concerned. If the Government asks the AAT to interpret a provision of a statute which is thoroughly vague and in which there seems to be no sensible policy underlying it, then of course it has to apply the techniques that a court would. It does not have anything else to go by. Whether the notion of giving weight to agency interpretations will work depends a great deal on how the Government creates the agency, how it defines its powers and what underpinning there is to those powers in the sense of a policy to guide the exercise of the powers. If there is no sensible policy to be discerned then, of course, the AAT is left in something of a quandary. It simply has to apply the techniques of interpretation that a court would apply.

Regarding the questions of ministerial responsibility, it is the case that Ministers have from time to time attacked the Tribunal for taking government policy out of the hands of the Government, where it is said to belong. That again is fair comment or not depending on the particular context. Again, if the legislation is framed in a way which confines the decision-maker and the AAT to operate within legal boundaries which do not make sense in terms of government policy, it is the Government's problem, not the Tribunal's, because the AAT must, as indeed a court must, stay within the bounds of the power described by the law. It seems to me that what the AAT can do, which would acknowledge ministerial responsibility to Parliament, is to give some weight to interpretations within those legal boundaries. The point about the new Senate procedures is that they will give a greater opportunity for Parliament to call officials to account for the policies that inform the application of statutory provisions. In that way, there should be greater scope for saying that ministerial responsibility, or indeed parliamentary supremacy, dictates some greater weight being accorded to administrative interpretations of the law.

Question:

Peter, you mentioned tenure at the end of your talk. Given the role that you see for the AAT, what do you think is an appropriate tenure for members of the Tribunal?

Response by Mr Bayne:

The question of tenure is one that has been considered by a select committee of Parliament , the Joint Select Committee on Tenure of Appointees to Commonwealth Tribunals¹⁹, and it has opted very much for giving members of what it called peak tribunals the same sort of tenure as a judge. It had very much in mind the senior members of the AAT and when it comes to the senior members of the AAT there is a strong case for saying there should be a strong form of tenure. But perhaps within the AAT there should be some differentiation of role. Those members who have 'judicial' status in terms of their tenure could avoid the run of the mill case and concentrate on cases which raise matters of principle; for example, what 'reasonable steps to attain work' means. Because the AAT's role is only indirectly impinging on policy, there is a case for saying that at least some of its members should have the degree of tenure that a judge should have so that, when they are called upon to make a difficult decision, there is a body of people there who are not fearful of their appointments not being renewed.

I should also add a perspective on tribunals which was not given much weight in the 1989 report of the Select Committee. It is a point of view that was put by Mr Ellicott to the Committee and it is that, if one sees tribunals as part of a system of administration of government and as having a major effect on policy-making through the constant and direct nature of their decisions, then one has to acknowledge that, when one considers questions of tenure, they are different to the courts. And just as the Government would not tolerate a senior public servant whose views on how powers should be exercised ran counter to government policy continuing to be in a position of influence in government, it seems to me that there has to be some capacity for Government to say that tribunals, which are doing essentially what administrators do, should also be capable of being restructured, and indeed abolished if they go off on a track which is contrary to policy.

This can be put in a somewhat more sophisticated way by saying that behind statutes are government policies. A statute is enacted with a viewpoint in mind that it will achieve a certain policy. Just as a new government may say that that policy is wrong or it has not worked and can repeal the statute, if there happens to be a particular tribunal with jurisdiction in that area then it may have to go as well if the policy which underpins the statute it administers is seen to be wrong-headed. Or it may be even that the particular tribunal is adopting a wrong-headed view of the policy.

That is a viewpoint, as I say, that was not accepted by the Select Committee. It is one that I would not apply to the AAT. The AAT, being a generalist tribunal, has a lot of jurisdictions and, of course, removing one jurisdiction rather than another can solve the problem if it is felt that the statute no longer makes sense. But, again, there may have to be some capacity to say to a member of a tribunal, 'You are just not doing your job in a way that is consistent with the policy of the statute'.

That does not mean that one removes judges on the same basis because judges' functions in relation to interpreting statutes are very different from those of the administrator. I am sympathetic to the views of former Senator Jim McClelland who was recently reported as saying that if the Government thinks that the Industrial Relations Commission does not have an independent role to play in relation to wage formulation, abolish the Tribunal: do not put people on it who are going to agree with your point of view and leave off those who are not.

So I think the question of tenure is a tricky one, but I think that what has been left out, or perhaps not given sufficient weight in the debate that took place after the Staples affair , which is what led to the Select Committee inquiry , is that the policy role of a tribunal was not seen clearly enough: it is a role which is very different to that of a court and which leads to a somewhat different answer to the question of tenure.

Question:

The question of the exercise of discretion and how departments resolve this is very much a matter of moment - particularly for, say, the Department of Immigration, the Department of Social Security and so on. There is a tendency to multiply rules so that we have some safe pen to fall back into. It seems to me that, whilst checks and balances are not part of the Westminster system, I think there must always be attention paid to achieving some social consensus. Whilst we have the separation of powers, we cannot dichotomise them into just the administration and the courts. Perhaps the tribunals have to make their own furrow. If they become subordinate, in a sense, to the Minister's policy then they are doing the very thing that the Administrative Decisions (Judicial Review) Act says that they should not do - that is, circumventing their discretion by direction.

Response by Mr Bayne:

I think that how one reconciles the suggestion that they give weight to administrative understandings of the law with the precepts of the ADJR is a matter that requires sophisticated argument. Where you have a tribunal you would not, of course, have an ADJR review - you would of course have an appeal on a question of law. But on a question of law the Federal Court would apply the principles stated in the ADJR Act.

It seems to me that the way to resolve it is to say that the principles that the Federal Court apply require agencies and the AAT to act within the boundaries of the law. But the court review should not go to the question of the merits of the exercise of the power within the law. It is in that area that giving weight to administrative understandings will not then cause the AAT to fall foul of the Federal Court on a question of law appeal. In other words, one has to acknowledge that there is an area of operation in the exercise of a discretion or the making of a discretionary judgment which is not the concern of the courts. So long as it is not the concern of the courts, what the AAT does is not going to lead to it being overruled on a question of law.

Lawyers working in this area know that it is hard to draw that line, but it is there and, despite the difficulty of doing it, it seems to me that the AAT can do as much as the administrator does. For example, take the question of unemployment benefits and taking steps to find work. The court is concerned to see whether the considerations which the agency has taken into account in making a judgment in the particular case are considerations relevant to the making of that judgment. I think it would be arguable that having regard to a person's retraining and the like is a relevant matter. And the court would not find anything unlawful in the Department saying that 'in general, we are not going to give credence to an argument that a person is looking for work by retraining her or himself or by running a small business'. If the agency says that, whilst at the same time not refusing altogether to look at the merits of a particular case, it is acting lawfully. And if the AAT was then to say, having regard to seeing it as part of a larger policy about the labour market: 'Well, we accept the validity of the agency's approach', it is not running the risk of being overruled by the court. The AAT should not, of course, yield legality control to the agency. But within the scope of the discretion or within the parameters of making the judgment, the AAT is not in any greater risk of its decision being overturned by the Federal Court than the agency is by having regard to some government policy. People will say: 'That is going to drive the AAT into a subordinate role'. It certainly qualifies the extent of its power. But it seems to me it does so in a way which is proper, having regard to its place in the system of government.

Let me make another point. The AAT is rarely concerned about how the agency has reached its decision. It may be apparent from the proceedings before the AAT that something has gone wrong in the way the agency has made its decision; for example, it may have acted upon obviously unreliable hearsay evidence and the like. But it is not a matter with which the AAT may be concerned because it will say: 'We make up our own mind about the facts'. It seems to me that the AAT might be concerned with the fact-finding procedures in agencies. To the extent that it improves those procedures, it has a normative effect, in the sense that its decision will affect cases other than those it is deciding upon.

So I would see in some respects a larger role for the AAT than the one it plays now. I do not think it is subordinate because I think that in the end the AAT is making up its own mind. Whether it gives weight to an administrative understanding of the law depends on a whole range of factors which I have not tried to articulate in this paper. It would simply take a lot more time to do that.

Question:

You say 'weight' but I seem to read that you say it is binding.

Response by Mr Bayne:

I have used the word 'weight' rather than the word 'binding' because I mean weight rather than binding. It may be that the AAT will say: 'Look, we understand that this interpretation is being taken, but we think it is simply not the right way to approach it'. But it would have to justify its view. It would have to then face the fact that it has less capacity to make that judgment in some situations. A situation in which it seems to me that the AAT could well have called for an explanation of policy, and then considered very carefully whether it should depart from it, was the situation of the application of the assets test under the Social Security Act and the Veterans' Affairs legislation. The AAT's approach has now been overruled by statute, which incidentally indicates the problem that it is undesirable that statutes should be constantly intervening to straighten out, if you like, from the Government's point of view, what the AAT does. At one point in applying the assets test - in making a judgment whether it was reasonable for someone to gain income from renting a property the value of which had been discounted for the purposes of a test - the AAT said, 'We will have regard to the tradition of passing properties from father to son.' That was a contentious view if you like. There was more than one point of view about that. Indeed, the legislative history of that provision would have shown that there was a debate, yet there was no account taken of it. The AAT should at least ask for an explanation, and if it is going to have its own policy, say clearly why that is preferable.

Question:

In which areas do you think the AAT has been most successful in dealing with the Social Security Act and which are its least successful areas?

Response by Mr Bayne:

I am sharpening the points in order to make an argument but there is no doubt that the AAT's role in hearing appeals under the Social Security Act has improved the quality of the administration. The manuals are better. They are written in a way which pays attention to AAT elaborations; in other words, they are much better guides than they used to be. I think that the administrators are now conscious of the fact that there is a review. I did a small amount of practice in this area before there was an AAT and it is just a different world altogether.

I can refer here to my earlier example of what it means to take reasonable steps to find work. At one point during the time I sat on the SSAT, I looked at the AAT cases to see what they said about that issue. I made my little summary, I looked at the manual and there it was in the manual. What had happened, I am not sure. Had the AAT looked at the manual or had the manual been rewritten after the AAT had had a look at it? Perhaps the latter. Perhaps that is an instance of where the AAT review, by articulating principles where there were none before, had a beneficial normative effect.

I think it has been least successful, for example, in the overpayments area where what has resulted is a stand-off between the Department and the AAT, although here it is the Federal Court that has articulated that different view rather than the AAT.

In relation to invalid pensions, there are obviously different points of view. The way the Act is worded makes it extremely difficult to do any more than guess in many cases. One of the problems is that the AAT members are guessing somewhat differently. There are different patterns of guesses around Australia and that is not particularly desirable.

That is the only way I can answer your question, but I do not want to suggest that the AAT's influence has not had the normative influence that it says it should have.

It concerns me that there is obviously a trend on the part of the Government to take discretions out of the Act and replace them with rules which really removes the whole point of the AAT articulating principles to guide the exercise of discretions. If it is now a matter of just applying the rule there is nothing left to explain other than what the rule is. There is no scope for having a more balanced judgment about what the outcome in a particular case should be. There is also the tendency to take tribunal jurisdiction away. It seems to me that that may have followed from what is the dominant approach of the AAT of not really looking at administrative policy when it is making up its own mind.

Even if it was just to say to the agency: 'What is it that you are trying to do? Tell us so that we understand', I think that that would lead to a healthier relationship. I think it should extend beyond the AAT hearing. I see no objection to the AAT members sitting down from time to time with the head of the department and saying: 'We are perceiving a problem in the way that your officers are deciding invalid pension cases. What's it all about?' You may also want to say to the welfare rights organisations: 'You had better come along too to see what we are saying'. That does not compromise independence. It will enable a better understanding to be developed. That is a lengthy answer to your question but the failures are perhaps less common than the successes. One objection you can make to my paper is: is it really such a big problem? I guess that is what you are asking. But it is a problem to a significant enough extent, it seems to me, to warrant analysis of it.