MASON C.J., WILSON, TOOHEY AND GAUDRON JJ. This is an appeal from a unanimous decision of the Full Court of the Federal Court of Australia (Northrop, Keely and Wilcox JJ.) which allowed an appeal from a decision of the Administrative Appeals Tribunal ("the Tribunal") and remitted the case to the Tribunal to be re-heard and determined according to law.

2. The respondent suffered personal injury in 1979 in the course of his employment by the Commonwealth. He received compensation in accordance with the provisions of the Compensation (Commonwealth Government Employees) Act 1971 (Cth) ("the Act"). In November 1982 he requested the Commissioner for Employees' Compensation ("the Commissioner") to redeem the Commonwealth's liability to make weekly payments for partial incapacity by payment of a lump sum. The request was made in accordance with s.49 of the Act. Initially, the Commissioner made a determination in favour of redemption but subsequently he revoked that determination. His final determination was that the respondent was totally incapacitated and that the request for redemption must therefore be refused. Section 49 permits redemption to take place only in circumstances where the worker is partially incapacitated for work and is not likely to become totally incapacitated (s.49(5)(a)).

3. The respondent applied to the Tribunal for the Commissioner's decision to be reviewed. It is necessary to recount the course of the proceedings in the Tribunal and the Federal Court. At the outset of the hearing before the Tribunal, the Commonwealth conceded, in the light of medical examinations of the respondent in 1984, that the Commissioner's determination that the respondent was totally incapacitated was erroneous and that there was no obstacle, based on the nature and extent of the incapacity, to the making of an order for redemption. The Tribunal, after resolving the further matters prerequisite to redemption described in pars.(b) and (c) of s.49(5) in a manner favourable to the respondent, determined that the liability of the Commonwealth should be redeemed and proceeded to determine the amount of the lump sum which was to be paid to the respondent. The parties were agreed as to the amount of the weekly payment due to the respondent at that time in respect of his partial incapacity and this sum formed the basis of the multiplicand upon which the computation of the lump sum was to proceed.

4. There were two principal issues upon which the parties were in contention at the hearing before the Tribunal. One was whether the Commonwealth was liable to continue to make weekly payments until the respondent's death (as the respondent contended) or only until he attained the age of sixty-five years (as the Commonwealth contended). The Tribunal resolved that issue in favour of the respondent and there is no appeal from that part of its decision. The other issue, and the one that was before the Federal Court, was whether a discount rate should be applied in calculating the multiplier and, if so, what the rate should be. Counsel for the respondent argued that if there was to be a discount then a rate of 2 per cent per annum should be adopted, this being the rate adopted by the High Court in Pennant Hills Restaurants Pty. Ltd. v. Barrell Insurances Pty. Ltd. [1981] HCA 3; (1981) 145 CLR 625.

5. In considering whether it was necessary to discount the lump sum, the Tribunal referred to the decision of this Court in Todorovic v. Waller [1981] HCA 72; (1981) 150 CLR 402 and continued:

view, unless there are such differences between the

common law personal injury cases and the payment of

a lump sum in redemption of liability to pay weekly

the High Court in Todorovic is inapplicable to such

redemption cases and found that, although significant for some purposes, they were of no substance in relation to the applicability of the reasoning in Todorovic and consequently that, in choosing a multiplier, a discount rate should be used.

6. The Tribunal then turned its attention to the choice of an appropriate rate. It had regard to the statement that was read by Gibbs C.J. immediately prior to the delivery of judgment in Todorovic. That statement reads (at p.409 of the report) as follows:

"Because of the practical importance of the

decision in these cases, the Court now publishes

this statement as to its effect.

In an action for damages for personal

injuries, evidence as to the likely course of

inflation, or of possible future changes in rates

of wages or of prices, is inadmissible. Where

there has been a loss of earning capacity which is

likely to lead to financial loss in the future, or

where the plaintiff's injuries will make it

necessary to expend in the future money to provide

medical or other services, or goods necessary for

the plaintiff's health or comfort, the present

value of the future loss ought to be quantified by

adopting a discount rate of 3 per cent in all

cases, subject, of course, to any relevant

statutory provisions. This rate is intended to

make the appropriate allowance for inflation, for

future changes in rates of wages generally or of

prices, and for tax (either actual or notional)

upon income from investment of the sum awarded. No

further allowance should be made for these

matters."

The Tribunal rejected the propriety of relying on Pennant Hills because that decision had been overtaken by Todorovic. It said that the decision in Todorovic was the result of a total reconsideration of matters relevant to discounting generally, and that the rate of 3 per cent per annum set therein must be used as a yardstick. The Tribunal went on to explain what it meant by the term "yardstick", saying:

"That does not mean that 3 per cent per annum is

necessarily the appropriate rate for such cases.

Our view is simply that, as the High Court fixed a

standard rate in common law personal injuries

cases, and for the same broad-brush reasons, a

fixed rate of discount should be applied in the

calculation of lump sums to be paid under section

49 of the Act where the liability being redeemed is

liability to make weekly payments for life or a

long period; and that the rate should bear an

appropriate relationship to the 3 per cent per

annum discount rate in common law personal injury

cases."

7. The Tribunal identified a distinction between the calculation required by Todorovic and that which was appropriate in s.49 cases. Todorovic had held 3 per cent per annum to be an appropriate discount rate in cases where the multiplicand is an amount nett of notional income tax. On the other hand, s.49(4) requires the amount of the lump sum to be "the value ... of the right of the employee to receive further payments of compensation under section 46". It is common ground that these words require the multiplicand in s.49 cases to be a gross amount determined without any reduction for notional income tax: cf. Watkins Ltd. v. Renata [1985] FCA 209; (1985) 8 FCR 65, at pp 72-73. Adopting the same broad-brush approach which this Court adopted in Todorovic, the Tribunal decided that 4.5 per cent per annum should be the rate of discount to be applied in calculating the lump sum payable to the respondent. It reasoned that this was probably the rate that would have been chosen in Todorovic if the multiplicand had been a gross amount without reduction for notional income tax. The matter was remitted to the Commissioner to make the necessary calculation.

8. The respondent appealed to the Full Court of the Federal Court. The grounds of appeal were quite specific. It was alleged that the Tribunal erred in law in finding that a rate of 4.5 per cent per annum was the appropriate rate, in failing to find that 2 per cent per annum or alternatively 3 per cent per annum was the appropriate rate, and in holding that there was any evidence to support its finding that 4.5 per cent per annum was the appropriate rate. There was no allegation that the Tribunal should have refused to adopt any discount rate.

9. The Full Court concluded that the Tribunal fell into legal error in regarding itself as bound by Todorovic to discount at all. The passage in the judgment of the Tribunal from where their Honours drew this conclusion has been set out earlier in these reasons. Notwithstanding the use of the word "bound", the passage may mean no more than that in the view of the Tribunal the reasoning of the majority in Todorovic impelled it to adopt a discount rate. Whether or not this be so is of no consequence now for the conclusion of the Federal Court to which we have referred was not germane to the outcome of the appeal to that Court. The parties were not in dispute over whether or not a discount rate should be chosen, but only over what the rate should be. The respondent (the appellant in the Full Court) was contending for a 2 per cent rate, or alternatively a 3 per cent rate in lieu of the Tribunal's determination of a 4.5 per cent rate. The Full Court discerned a question of law which underlay this disputation between the parties, namely whether the Tribunal fell into error in selecting a discount figure "by preferring to the evidence before it a divination of the likely position, upon a matter of fact, of the members of the High Court". Whether there was in truth a question of law is a matter we do not have to decide. The principal issue in the appeal to this Court is whether the Full Court was correct in giving an affirmative answer to the question itself.

10. The Full Court allowed the appeal, set aside the decision of the Tribunal and remitted the matter to the Tribunal for re-hearing. The Court envisaged that further evidence would be received. In order to assist the Tribunal in its task, the Court made some observations as to the legal principles which ought to underlie the re-assessment including, inter alia, the following:

"It can only be said that there is a value in

receiving payment in a present lump sum rather than

in wage-rate adjusted weekly amounts where it can

be said that the level of interest rates is likely

... to exceed the level of wage increases; that is,

comparing interest rates to increases in wage rates

-- not prices -- expressed in percentage terms,

that there is a positive rate of real interest.

Actual rates being impossible to predict, this

conclusion can only be reached if, in a particular

case, there is material to show that there has in

the past existed a relationship between interest

rates and increases in wage rates sufficiently

constant to justify the assumption that a given

relationship will apply in the future. If such

material is not available, there is no basis for

applying any discount at all."

11. In our opinion this approach cannot be supported. It fails to recognize the true significance of this Court's decision in Todorovic. It was far more than a decision of fact based on the evidence adduced in that particular case. It was a decision which took the unusual course of prescribing a rule of practice for future cases. This extraordinary course was prompted by a recognition of the magnitude of the difficulty that confronted courts as they sought to provide fair and just compensation to plaintiffs in personal injury cases in respect of losses to be suffered far into the future, and by the importance of predictability in the assessment of damages. A majority of the Court was satisfied that a plaintiff obtains an advantage when receiving present payment of a sum of money which in other circumstances would not be received until a future date (Gibbs C.J. and Wilson J. at pp.413-414; Mason J. at pp.442-443; Aickin J. at p.460; Brennan J. at pp.466-467, 477-478). That conclusion was reached notwithstanding evidence of extraordinary fluctuations during the seventies in both inflation and interest rates. The settled doctrine which had required courts to reject evidence of future inflation was adhered to because it would be purely speculative, would tend to prolong trials and would introduce an additional element of uncertainty into awards. It was recognized that some allowance must be made to counter the effect of tax on the income produced by investment of the lump sum but that the speculative elements in such a task defied any pretence at precision. Gibbs C.J. and Wilson J. said, at p.423:

"No assumption can safely be made as to what the

tax scales will be in the near future, still less a

decade hence. For these reasons, notional tax can

be taken into account only in the broadest way, by

making an adjustment to the discount rate, and it

is impossible to make even a pretence of accuracy

in doing so."

Mason J., at p.449, posed the question whether an attempt should be made to calculate and allow for tax in each particular case and rejected it in favour of a rule of thumb approach which fixed a lower discount rate to be applied in every case, whatever the plaintiff's liability to tax may be. He favoured this approach

"partly because we cannot hope to achieve precision

in the assessment of damages and partly because

simplification of the assessment of damages is very

desirable."

12. The preliminary statement which accompanied the Court's decision in Todorovic made it clear that the discount rate of 3 per cent per annum was intended to make the appropriate allowance for inflation, for future changes in rates of wages generally or of prices, and for tax upon income from investment of the sum awarded. Subject to any relevant statutory provisions, it was to be applied by all courts in the assessment of damages until such time as there was sufficient change in general economic circumstances to justify the Court in undertaking a review.

13. As we have seen, the Tribunal considered that notwithstanding the distinctions that could be drawn between personal injury cases and redemption cases, those distinctions did not render the reasoning in Todorovic to be inapplicable to s.49 cases. It therefore determined that, for the same broad-brush reasons, a fixed rate of discount should be applied in the quantification of a redemption sum where a long period was involved.

14. In our opinion, this general approach of the Tribunal was correct. Just as it was thought in Todorovic to be desirable to relieve courts from the complexity and diversity of result involved in each court exercising its own discretion in the choice of a discount rate, so it is desirable in principle to facilitate the task of the Commissioner in determining the amount of a lump sum by way of redemption. There is no justification for rejecting the reasoning in Todorovic as having no application to the Commissioner's task and thereby to leave the Commissioner, and on review the Tribunal, in the same predicament that confronted the courts prior to Todorovic. This is so notwithstanding that s.25(9) and (11) of the Act requires expressly that regard be had to wage increases that the worker might have expected to receive if he had not been incapacitated. That requirement is no different in principle from the obligation of a court, in adopting a multiplicand to form the basis of an assessment of damages for future economic loss, to have regard to the increased wages which the plaintiff might have expected to receive. It is simply not feasible, as the Full Court did, to require the Commissioner and the Tribunal to have regard to material to show that "there has in the past existed a relationship between interest rates and increases in wage rates sufficiently constant to justify the assumption that a given relationship will apply in the future" and that in the absence of such material there should be a nil discount. Todorovic supplies a different approach which in our view is wholly compatible with the requirements of the Act.

15. This Court was invited in 1986 to review the general economic situation with a view to reconsidering what the discount rate, if any, should now be; but the application for special leave to appeal which served as the vehicle for that invitation was refused for the reason that a general review was not warranted at that stage: Ioannou v. Mitchell, Melbourne, 8 August 1986.

16. It follows that we are unable to agree with the Full Court's answer to the second question. The Tribunal did not fall into legal error by adopting, whether or not it considered itself to be bound to do so, the course proposed by this Court in Todorovic. Nevertheless, we think that the Tribunal misdirected itself in arriving at a discount rate of 4.5 per cent per annum. The fact that the quantification of the loss of earning capacity for which the plaintiff in a personal injury case is to be compensated leads to the adoption of a figure which is nett of income tax, does not mean that the chosen discount rate of 3 per cent per annum is inappropriate to a case where what is to be valued for the purposes of arriving at a redemption figure under the Act is "the value ... of the right of the employee to receive further payments of compensation". The valuation of that right is a process to which the fact that had the employee received weekly payments in the future he would have been subject to income tax in respect of those payments is irrelevant. The Act is not concerned with what might have happened after the employee received such payments. The employee is entitled to receive in his hands by way of redemption a sum which represents the present value of the gross sum of the weekly payments which in the absence of redemption he might have expected to receive. On the other hand, the plaintiff in a personal injury case is only entitled to compensation for what he has lost. The assessment of that loss requires the application of a different criterion. We think, therefore, that there could be no objection to the Tribunal, in the exercise of its discretion, choosing to adopt a discount rate of 3 per cent per annum in the present case.

17. The appeal should be allowed. The order of the Full Court should be varied so as to remit the matter to the Tribunal, not with a view to a re-hearing but to enable the Tribunal to reconsider the discount rate per annum which the Commissioner should use in making his computation.

DEANE J. In Todorovic v. Waller [1981] HCA 72; (1981) 150 CLR 402, this Court was called upon to consider what, if any, rate of discount should be applied by Australian courts to the sum of estimated lost future net (i.e. after tax) earnings to ascertain the appropriate amount to compensate for loss of future earning capacity in an action for damages for personal injury. The effect of the order which was made with the concurrence of a majority of the whole Court was summarized in the statement published by Gibbs C.J. at the time of delivery of the individual judgments (at p.409):

"Where there has been a loss of earning capacity

which is likely to lead to financial loss in the

future, or where the plaintiff's injuries will make

it necessary to expend in the future money to

provide medical or other services, or goods

necessary for the plaintiff's health or comfort,

the present value of the future loss ought to be

quantified by adopting a discount rate of 3 per

cent in all cases, subject, of course, to any

relevant statutory provisions. This rate is

intended to make the appropriate allowance for

inflation, for future changes in rates of wages

generally or of prices, and for tax (either actual

or notional) upon income from investment of the sum

awarded. No further allowance should be made for

these matters."

2. As that statement made plain, Todorovic involved the assumption by the Court of what was essentially a legislative function. The identification of a general discount rate of 3 percent represented the prescription of a rule of practice for the future rather than the elucidation of a principle of law or the ascertainment of existing facts (cf. per Holmes J., Prentis v. Atlantic Coast Line Co. [1908] USSC 160; (1908) 211 US 210, at p 226; and per Brennan J., Todorovic, at pp 464-470). Indeed, the precise orders made were not supported by the judgments of a majority of the justices who joined in making them (Gibbs C.J., Mason, Aickin, Wilson and Brennan JJ.; Stephen and Murphy JJ. dissenting). Of those justices comprising the majority, Gibbs C.J., Wilson J. and (semble) Aickin J. favoured a higher discount rate (4 percent); Mason J. favoured a lower rate (2 percent); while Brennan J. accepted a discount rate of 3 percent as falling within an acceptable range. The extent of the differences of opinion between the majority justices and of the compromise involved in the adoption of the Todorovic orders can only be fully appreciated in the light of practical example. Thus, if one takes an award of damages to compensate a person for loss of net earnings of $300 per week over a future estimated working period of thirty years, the 2 percent discount favoured by Mason J. would result in an award of approximately $354,000; the 3 percent under the Todorovic order would result in an award of approximately $311,000; while the 4 percent favoured by Gibbs C.J., Aickin and Wilson JJ. would result in an award of approximately $276,000.

3. To draw attention to those unusual aspects of the decision in Todorovic is not to criticize it. The detailed judgments in the case display a full appreciation of the difficulties and unusual features of the task which the Court was called upon reluctantly to essay. The performance of that task was unavoidable if there were to be consistency in the administration of justice in cases involving the award of damages for personal injuries. The relevance of those unusual aspects of Todorovic for present purposes is, however, plain enough. The case was a judicial decision in so far as it resolved the particular litigation before the Court. Beyond that, the decision in the case cannot properly be understood as constituting judicial authority for the proposition that, even for the purposes of curial proceedings for damages for personal injury, 3 percent represented the correct discount rate. That was "essentially a matter for judgment, and different minds would attribute a different weight to the immeasurable factors in selecting a discount rate" (per Brennan J., at p 478 and see also the reference to Hawkins v. Lindsley in the judgment of Gibbs C.J. and Wilson J. at p 423). The authority of the decision for future cases lay in the fact that, in practice, the discount rate adopted by the Court would be "its guide in determining appeals challenging the assessment of damages", and would thus be a rate "effective to guide the courts from which those appeals" might come (per Brennan J., at p.478). Otherwise, the discount rate selected by the Court lacked the authority of a decision of the Court on a question of law. Clearly, the Court's prescription of a discount rate of 3 percent to be applied by courts in assessing damages for personal injury was not directed to, or binding upon, a statutory office holder who might be required, in the performance of an administrative function entrusted to him, to determine what, if any, discount rate should be applied for the purposes of an administrative valuation of a right to receive future payments of money over a period of years. Apart from the requirements of procedural fairness, such statutory office holders are not restricted by the limits and restraints of judicial process and procedure. They may possess expert qualifications and knowledge not ordinarily possessed by a judge. They may properly have regard to considerations and factors which have been seen as too speculative or uncertain to be the subject of judicial determination. In performing their administrative functions, they are not ordinarily subject to being controlled in advance or subsequently overridden by the courts otherwise than on questions of law.

4. The function performed by the Administrative Appeals Tribunal ("the Tribunal") in the present case was an administrative one which was entrusted to the Commissioner for Employees' Compensation ("the Commissioner") by s.49(3) and (4) of the Compensation (Commonwealth Government Employees) Act 1971 (Cth) ("the Act"). That function was the "determination" of the present "value" of "the right of (the respondent) to receive further payments of compensation under section 46" of the Act. In performing that function, the members of the Tribunal did not act as a court. They stood in the shoes of the Commissioner. Nonetheless, they regarded themselves as "bound" by Todorovic to determine the value of the respondent's right by applying a "discount rate" which, they concluded, "must be set by using as a yardstick the rate of 3 percent per annum set in Todorovic." It follows from what has been said above that the members of the Tribunal fell into error of law in regarding themselves as bound by Todorovic to apply any such discount rate. As has been seen, Todorovic is simply not applicable, either as judicial authority or as legislative prescription, to control administrative office holders in the performance of their administrative functions. Moreover, even if Todorovic were applicable to control the performance of administrative functions, the differences between the function of the trial court in Todorovic and the function of the Tribunal in the present case were such that Todorovic could not properly be seen as establishing a precedent or "yardstick" which the Tribunal was bound to follow or adopt.

5. The administrative determination under s.49(3) and (4) of the Act of the amount of the lump sum which represents the present value of the right of an employee to receive further payments of compensation differs markedly, in several respects, from the judicial determination of the appropriate lump sum amount to compensate an injured worker for loss of earning capacity. First, once life expectancy has been determined, the right of an employee under the Act is not subject to the uncertainties which support discounting in the case of the ordinary action for damages for loss of earning capacity. There is no risk of loss of employment for other reasons to be taken into account. There is no danger that the employer might have become insolvent or otherwise gone out of business. There is no need to make allowance for the possibility of incapacity from other causes. All those matters of uncertainty were present and relevant in Todorovic as factors supporting a positive discount rate. Indeed, it is arguable that, in the case of at least one of the majority judgments, they were of critical importance (see the reasons given by Brennan J. (at p.475) for concluding that 2 percent should be added to the Government bond rate in identifying an appropriate return on investments). They have no part to play in a valuation of the accrued right of an employee to receive future payments of compensation from the Commonwealth under the Act. Secondly, the right being valued pursuant to s.49(3) and (4) is a right to receive payments under a scheme which provides for automatic adjustment in accordance with variations in, among other things, award rates (Act, s.25(9) and (11)) which can ordinarily be expected, at least over the long term, to reflect the effects of inflation. That being so, one possible starting point of an administrative valuation under s.49(3) and (4) of the Act would be recognition of the fact that what is being valued is the right to receive payments which will periodically increase for so long as inflation remains. The Tribunal was obviously not bound to adopt and apply, for the purposes of its administrative function, the approach which had been accepted in this and other courts as appropriate to judicial proceedings, namely, that direct account cannot be taken of the likely effects of future rates of inflation for the reason that they are not susceptible of judicial determination. The legislative prescription in Todorovic of a compromise discount rate of 3 percent to be applied by courts in assessing damages for the loss of future net earnings was justified, in the context of the need for judicial consistency, by a combination of factors arising to a significant extent from the limitations of judicial expertise and procedures. There is no hint in any of the judgments in that case that the Court was purporting to create judicial authority or legislative prescription binding an administrator entrusted with a statutory task of assessing the present value of an accrued right to receive future payments from the Commonwealth to adopt a particular procedure or discount rate in the performance of that statutory task. Nor was there any basis or occasion in law or in fact for the creation by the Court of such a judicial authority or legislative prescription. That being so, it was open to the Tribunal, if it thought it appropriate, to adopt the approach of seeking, in the light of accumulated knowledge and available expertise, to reach an informed view about the likely extent of future inflation and to devise a formula which took more direct account of inflation as a positive factor which would inevitably increase the amounts of compensation which would become payable under the Act. On the other hand, it was open to the Tribunal to adopt, as was done by this Court in Todorovic, the indirect "broad brush" approach of taking account of future inflation only as a factor in estimating likely "real rates" of future interest. If the Tribunal concluded that the preferable approach to the performance of its administrative function was the former one ("the direct approach"), Todorovic did not compel it to disregard its own view and adopt the latter one. Thirdly, even if the approach were to be taken by the Commissioner or the Tribunal that rates of future inflation were so uncertain that inflation should be taken into account only indirectly in determining a "real rate" of interest, the evidence before the Tribunal in the present case provided strong support, even in the context of Todorovic itself, for a rejection of the rate of 3 percent as a starting point or yardstick. I shall endeavour to explain why that was so.

6. The evidence before the Tribunal (see report of Mr. N.D. Whitehead from the Office of Australian Government Actuary) was that over the whole period from 1950 to 1981 the average "real" rate of interest (i.e. actual available rate of interest less rate of inflation) was 1 percent if Government bonds were taken as the appropriate reference point. As Mason J. pointed out in Todorovic (at p.448), this "real" rate is, to some extent, distorted by the effect of Government financial policies during some years of high inflation. The fact remains, however, that during two of the last three completed decades, the average "real" rate of interest on Government bonds was less than 2 percent; in the other decade, 1960-1969, the average "real" rate was 2.5 percent. The actual rate of interest available from investment in Government bonds is traditionally a comparatively low one by reason of the absence of risk of non-payment of interest or loss of capital. It is, however, at least arguable that, life expectancy having been determined and taken into account, it was an appropriate reference point in valuing a vested right to receive future payments from the Government since the solvency and reliability of the payer was identical in each case (cf. Todorovic, at p.475). The "real" rate of interest to which Mr. Whitehead referred did not, however, allow for the fact that interest received constitutes assessable income for taxation purposes. In that regard, as was pointed out in judgments in Todorovic (see, e.g., per Stephen J., at p.435 and per Mason J., at pp.443-444), it is essential to bear in mind that tax is effectively payable on the amount of interest received without a deduction being available to counterbalance the component of the interest which is realistically to be seen as compensating for the effects of inflation upon the true value of the capital lent. The importance of this can be best demonstrated by an example which is not completely hypothetical. If, for example, the Government bond rate was 12.03 percent and the rate of inflation 7.04 percent over a particular period, the "real" rate of interest, if income tax be ignored, would be a positive 4.99 percent. Even ignoring the more precise effect of differing rates of inflation (see, e.g., the formula in Rea, "Inflation, Taxation and Damage Assessment", Canadian Bar Review, vol.58 (1980) 280, at p.283 which would provide a rate of 4.66 percent in the postulated example), that "real" rate is misleading in the "real" world of income tax since no deduction for inflation is allowable from the gross interest in determining taxable income. The consequence is that the amount of interest retained after tax would, in the postulated example and assuming a rate of tax of 25 percent, be 9.02 percent. It follows that, assuming a rate of tax which is somewhat unfavourable from the employee's point of view (the expert witnesses before the Tribunal suggested the appropriate notional rate of income tax was 30 percent), the "real" rate of interest, after allowance for tax, would be less than 2 percent. On this approach, the effect of the expert evidence from the office of the Government Actuary is that, if allowance is made for tax, the average "real" Government bond rate of interest over the years 1950-1981, and in two of the last three completed decades, was a negative one. It may well be thought appropriate by an administrative decision maker such as the Commissioner or the Tribunal to adjust traditional "real" rates of interest so as to afford additional weight to contemporary rates. Even if that be done, however, it is difficult to discern, in the absence of any compensatory factor, convincing justification of a discount rate of 3 percent as an objectively fair one if regard is to be had to the full effect of taxation upon the proceeds of loan investments and if the rate of interest paid by the Government itself be accepted as an appropriate reference point for determining the value of an invested right to receive payments from the Government. In Todorovic, there was an inbuilt (albeit inadequate) compensatory factor for the effect of income tax upon the inflation component of the gross interest which would be earned by invested capital in that, notwithstanding that what was involved was compensation for the loss of net (i.e. after tax) earnings, no allowance was made for the effect of a progressive and unindexed taxation scale upon inflated earnings (see per Stephen J., at p.434, per Brennan J., at p.476). No such compensatory factor exists in the present case however since the multiplicand to which the members of the Tribunal applied the discount rate represented future gross payments of compensation upon which varying tax rates would not, in any event, have any effect. In these circumstances and in the absence of some other compensatory factor, the importance of making fair allowance in any discount rate for the effect of tax upon the inflation component of investment income cannot be gainsaid. There are two points which should be made, for present purposes, about the need to make such fair allowance in the circumstances of the present case. The first is that the evidence before the Tribunal plainly indicated that, at least in the absence of a compensatory factor of a kind which existed in Todorovic, the adoption of the Todorovic rate of 3 percent would fail to make it. The second is that the making of such a fair allowance otherwise than by some "broad brush" compensatory factor would seem to involve a need for the administrator to enter into areas of assumption and assessment (future rates of tax and inflation) which are peculiarly his own in that they go beyond the permissible primary function of a court under the Todorovic formulation.

7. The effect of what has been said above may be summarized as follows. It was open to the Tribunal to determine the present value of the respondent's right to receive payments of compensation in a manner which made no allowance at all for a discount rate of the type applied by this Court in Todorovic. The respondent's "right" in the present case was to be paid compensation under a scheme which contained inbuilt safeguards (Act, s.25(9) and (11)) to ensure that the amount of compensation did not remain constant in the face of increases in real wages. The present value of that right to receive future compensation must inevitably be affected by expectations as to future rates of interest, inflation and income tax. Todorovic does not stand as authority for the proposition that administrative office holders, such as the Commissioner or the members of the Tribunal, are bound, in discharging a statutory function of valuing such a right, to take as a starting point the total of the future payments of compensation which would be received by the employee on the unreal and (from the employee's point of view) unfavourable hypothesis that the amount of those future payments would remain constant. Even if that hypothetical total be taken as a convenient starting point, Todorovic does not stand as authority for the proposition that that total must be further reduced adversely to the employee. In particular, it does not require that the total be further reduced by the application of the Todorovic discount rate or by some other discount rate derived from it. The Tribunal was, of course, entitled to follow the lead of this Court in Todorovic and, in the interests of consistency, set a general procedure to be followed by the Commissioner in assessing the present value of the right to receive future payments of compensation. It may have been open to the Tribunal to conclude that, notwithstanding obvious differences and harshness from an employee's point of view, the Todorovic discount rate was the appropriate one. What was not open to the Tribunal was to see Todorovic as laying down some binding principle which obliged it to accept and adopt the Todorovic procedure and to apply, as a compulsory yardstick, the compromise Todorovic discount rate of 3 percent.

8. In these circumstances, the members of the Full Court of the Federal Court were correct in concluding that the decision of the Tribunal should be set aside and that the case should be remitted to the Tribunal to be heard and decided again, with such further evidence as the Tribunal may admit, according to law. In that regard, it should be mentioned that it emerged during argument that it was common ground before the Tribunal in the present case that some rate of discount should be applied by the Tribunal to the total of future payments of compensation calculated on the hypothesis that those future payments would remain constant. However, in view of the importance of the matters involved and of the fact that the matter must be remitted to the Tribunal in any event, the Tribunal should not be precluded from examining for itself the question of the appropriate valuation procedure to be followed, the question whether it was necessary to adopt any rate of discount, and the question whether, if it was, the appropriate discount rate in the present and foreseeable economic circumstances was positive, negative or zero.

9. There remains to be mentioned the question whether the Tribunal was entitled to pay regard to the effect of income tax upon the future payments of compensation which the respondent would be expected to receive in the absence of redemption. The Tribunal added 1.5 percent to the Todorovic rate to take account of the effect of such income tax. I agree with Mason C.J., Wilson, Toohey and Gaudron JJ. that the respondent was entitled to receive by way of redemption a sum which represented the present value of the gross sum of the weekly payments which, in the absence of redemption, he might have expected to receive. Once the Commissioner has determined that the liability to make further payments is to be redeemed (cf. s.49(5)(c)), the Act is not concerned with what might have happened after the respondent received such payments. That being so, it was not the task of the Tribunal to assess how much of the fruits of the right to receive future payments of compensation the respondent would have been able to retain. On the other hand, in valuing the right of the respondent to receive future payments of compensation, the Tribunal was, in my view, entitled to pay regard to the fact that the objective market value of that right, as a right to receive future payments of income, would be affected by the liability to income tax which receipts of income in excess of the tax threshold prima facie attract in that it is a fact of the market place that a right to receive capital will, in the context of taxation laws, ordinarily have a higher value than a right to receive a corresponding amount of income. The question of what allowance should be made is a matter for the Tribunal upon which its decision is final and binding unless affected by error of law. I certainly do not possess the expertise or knowledge to express a view, let alone to purport to pronounce a binding principle of law, in that regard.

Appeal allowed. The appellant to pay the respondet's costs of the appeal.

Order that the order of the Full Court of the Federal Court dated 18 July 1986 be varied by deleting paragraph 2 and in lieu thereof order as follows:

2. Remit the matter to the Administrative Appeals Tribunal

to enable it to reconsider paragraph 1(d) of its decision of

26 April 1985 in respect of the discount rate per annum which

the Commissioner for Employees' Compensation should use in

making his determination of the lump sum payable to the

applicant under s.49 of the Compensation (Commonwealth

Government Employees) Act 1971 (Cth).