by Jonathan Gaunt QC
In 1982 George Grover and I gave a paper in this series entitled Revolution in the arbitrator’s duties(1). It examined the effect on rent review arbitrations of the 1979 Act and the guidelines laid down for the grant of leave to appeal in The Nema.
In my paper I asked whether the then celebrated(2) rent review valuation cases decided under the case stated procedure would have got so far as the courts under the 1979 regime. I concluded that, in future, it would be a very rare case in which the parties would succeed in obtaining a ruling of the court upon the construction of the valuation provisions of rent review clauses which provided for arbitration. It was implicit in that conclusion, but unstated in my paper, that parties would not simply bypass the Act by issuing construction summonses, because, I must have thought, the other side would apply for a stay under section 4 of the 1950 Act.
Second, I put forward three reasons why the Nema guidelines (Pioneer Shipping v BTP Tioxide Ltd [2] AC 794) might have to be modified so as to deal with rent reviews:
(i) in rent reviews arbitration is not chosen because speed is of the essence: the clauses often expressly provide for late determination and a determination by an expert would be quicker;
(ii) rent review clauses may or may not involve standard clauses but would tend to involve standard elements;
(iii) while the facts would almost inevitably be “one off”, points of law would often not be “one off” because there would be further reviews on the same basis and the parties would be bound on later reviews by the decision of the arbitrator on an issue in an earlier review by reason of the doctrine of issue estoppel.
Third, I hazarded a guess that one might see an increase in the popularity of the speaking expert, that is an expert contractually obliged to give his reasons, because it appeared to be no more difficult to impugn his decision, as the law then stood, than that of an arbitrator. Fourth, I predicted an increase in attempts to get awards set aside for misconduct. Fifth, I thought that the criteria for the determination of a preliminary issue under section 2 would rarely be satisfied, certainly not once the majority of the expert witnesses’ fees had been incurred.
It is interesting to inquire now how far those various predictions have proved to be right. I have analysed the cases concerning the construction of the valuation elements of rent review clauses which have been reported in EGLR during the period 1985 to 1990, omitting appellate cases (to avoid double-counting). The analysis shows that parties to rent review clauses who anticipate a legal difficulty simply bypass the 1979 Act by applying to the court by way of originating summons for a declaration as to the true construction of the clause. Nobody ever seems to apply for a stay of such proceedings under section 4 of the 1950 Act as they would be entitled to do. Of such cases, 43 have been reported in the period in question (some of which, admittedly, will have involved experts rather than arbitrators). In the same period only six appeals have been reported and of these only two have been allowed. The first was National Westminster Bank plc v Arthur Young McClelland Moores & Co [5] 1 EGLR 61; (1984) 273 EG 402, which we now know was wrongly decided and should never have got leave. The other was Electricity Supply Nominees Ltd v London Clubs Ltd [1988] 2 EGLR 152; [1988] 34 EG 71 which was an appeal by consent. In the same period six preliminary issues under section 2 of the Act are reported and nine attempts to attack an award for misconduct (of which five were successful). Eleven applications for leave to appeal are reported, of which only three were granted. These applications do not overlap the six decided appeals, which must be added to the total of reported cases where leave was granted.
It would be unfair to Estates Gazette to regard the cases which it reports as totally random and thus perhaps a representative sample; nevertheless one would expect it to be more likely than not to report successful appeals on points of law connected with valuation and so the above exercise, while obviously not scientific, does none the less provide an indication of what is going on. It points strongly to the conclusion that parties prefer to sort out legal difficulties before going to an arbitrator and that the risk of the rent review arbitrator being successfully appealed is almost negligible; he is more likely to be vulnerable to criticism on the ground of misconduct or procedural mishap.
Experts
Shortly after I gave my 1982 paper, the courts appeared to give considerable support to the view that there was a significant difference between speaking and non-speaking awards by experts, such that a speaking award could be declared a nullity if found to be based on a wrong principle: see Burgess v Purchase & Sons (Farms) Ltd [3] Ch 216. Such a wrong principle, so the argument went, might include a wrong principle of valuation based on a misinterpretation of the rent review clause: see eg Apus Properties Ltd v Douglas Farrow & Co Ltd [1989] 2 EGLR 265. This view has recently been wholly discredited by the decision of the Court of Appeal in Jones v Sherwood Computers [1989] EGCS 172 (otherwise unreported) applied by Knox J in a rent review context in Nikko Hotels (UK) Ltd v MEPC plc [1991] NPC 41; see also this week’s issue at p 86. In the latter case it was argued for the tenant that the expert had misconstrued “the average room rate” as defined in the lease and therefore awarded the wrong figure. Knox J held that so long as the expert answered the right question, albeit wrongly, his decision was binding on the parties. The question of construction was one of the questions referred to the expert; in those circumstances the court did not need to inquire whether his interpretation of the lease was right or not. In practical terms this means that experts are totally unappealable and it follows that one way of making a rent review award judge proof is to provide for such disputes to be referred to an expert rather than an arbitrator.
Exclusion agreements
It is open to the parties to an arbitration, after the commencement of the arbitration, to make an agreement excluding the right of appeal. The effect of such an agreement is to preclude the High Court from (a) granting leave to appeal or (b) ordering the arbitrator to state reasons or (c) entertaining an application for the determination of a preliminary issue. I have not come across a case in which parties to a rent review have made such an agreement and it is perhaps difficult to see what advantage there would be in doing so at that stage.
Appeals
The subsequent history of The Nema guidelines is well known. There seems to have been a tendency for judges to give leave to appeal in any case involving a mildly intriguing point of law and so the House of Lords saw fit in Antaios Compania Naviera SA v Salen Rederierna AB [5] AC 191 to reaffirm the guidelines in these trenchant terms at p 206D:
…all that the judge has to decide on the application is: first, is this dispute, on the one hand, about a one-off clause or event, or, on the other hand, about a standard term or an event which is a common occurrence in the trade or commercial activity concerned? If it is the former, he must then consider: whether the arbitrator was in the judge’s view so obviously wrong as to preclude the possibility that he might be right; if it is the latter he must then consider whether a strong prima facie case has been made out that the arbitrator was wrong? Unless the answer he would give to the question appropriate to the type of case to which the application with which he is concerned is: “Yes,” he should refuse leave to appeal.
Moreover, their lordships indicated that reasons should not be given for the grant or refusal of leave save in a case where the judge was minded to grant leave to appeal to the Court of Appeal against his decision, which he should do only if there are conflicting decisions (as opposed to dicta) on the point of law in question or other good reasons why English commercial law would be clarified by the matter being considered by the Court of Appeal.
After that we waited with bated breath for a decision upon the question whether the Nema guidelines required to be modified in a rent review case. In Lucas Industries plc v Welsh Development Agency [6] 1 EGLR 147; (1986) 278 EG 878 the Vice-Chancellor held that they did for much the same reasons which I had advanced in my 1982 paper. He concluded that the right test should be whether the judge was left in real doubt as to whether the arbitrator was right in law; if so, he should grant leave to appeal. This approach has since been disapproved by the Court of Appeal in Ipswich Borough Council v Fisons plc [1990] 1 EGLR 17; [1990] 04 EG 127, which introduces the concept of a continuum of suspicion. There is a presumption of finality, the rebuttal of which requires to be based on a suspicion that the arbitrator was wrong, the degree of suspicion varying with the particular circumstances of the individual case according to whether the contract and the effect are one off (in which case obvious error has to be shown): or whether the dispute involves a standard term and an effect which will be repeated (in which case a more or less strong prima facie case of error must be shown). One may be forgiven for wondering what practical difference there is between “suspicion of error” and “real doubt as to correctness” as a test for the grant of leave; probably there is none in a standard term/repeated event case; both tests produced the same result in the Ipswich case; the lesson of that decision is that in a one-off case the test is stricter.
Indeed one may be forgiven for thinking that the debate as to the correct semantic formulation of the test for granting leave is arid and academic. The fact of the matter is that leave is very rarely being granted and appeals are almost never allowed. I conclude that the combined effect of the 1979 Act and the efforts of the gentlemen appointed by the president of the RICS are that rent review awards are almost always in practice judge proof and that both Parliament and the RICS can take joint credit for having achieved this.
Moreover, the history of the 1979 Act in the context of rent review provides a salutary lesson for judges. In the only reported case where leave was granted and the appeal allowed, namely National Westminster Bank v Arthur Young, the judge got it wrong. He did so because he applied a mechanistic verbal analysis to a problem to which the commercial common-sense answer was obvious. The case is a copybook example of the situation which Lord Diplock had in mind in the Antaios when he said at p 210D:
If detailed semantic and syntactic analysis of words in a commercial contract is going to lead to a conclusion that flouts business common sense, it must be made to yield to business common sense.
In his own field the arbitrator is better placed to appreciate the business context than a judge. Valuers, after all, know more about valuation than lawyers. The valuer arbitrator is, if anything, more likely to get the answer right.
Misconduct and procedural mishap
Under section 22 of the 1950 Act the High Court has power to remit any of the matters referred to arbitration for the reconsideration of the arbitrator and under section 23 to set aside an award where an arbitrator has misconducted himself or the proceedings. The following behaviour has been held to justify the remission of the award:
(i) failure to copy one party’s communications to the arbitrator to the other side: Shield Properties & Investments Ltd v Anglo Overseas Transport Co Ltd [5] 1 EGLR 7; (1984) 273 EG 69;
(ii) admission of inadmissible evidence if objected to in time: Townsend Securities Ltd v William Morrison Supermarkets [2] 1 RR 239;
(iii) failing to enforce a direction that comparables should be within the direct knowledge of the witness or verified by documentary evidence; omitting to forward a letter to the other side; failing, contrary to a prior direction, to raise with the parties whether they wanted an oral hearing: Control Securities plc v Spencer [9] 1 EGLR 136; [1989] 07 EG 82;
(iv) taking into account material obtained from the arbitrator’s own knowledge or research without giving the parties an opportunity to deal with it: Top Shop Estates Ltd v Danino [5] 1 EGLR 9; (1984) 273 EG 197;
(v) rejecting the expert evidence of one party on a ground not argued or adverted to in the arbitration and on the basis of facts known to the arbitrator, without putting those matters to the party in question and giving him a chance to deal with them: Fisher v P G Wellfair Ltd (1982) 263 EG 589 and 657;
(vi) rejecting a zoning analysis applied by both parties’ surveyors without giving either an opportunity to express a view on the arbitrator’s rejection of such an approach: Zermalt Holdings SA v Nu-Life Upholstery Repairs Ltd [5] 2 EGLR 14; (1985) 275 EG 1134;
(vii) failing to send a copy of each party’s cross-representations to the other: Banner Industrial & Commercial Properties Ltd v Clark Paterson Ltd [0] 2 EGLR 139; [1990] 47 EG 64.
In commercial cases, applications to set aside for misconduct or remit are in fact quite common. In the law reports proper for the period 1981 to 1989 there are no less than 43 such cases reported. These cases show:
(a) that to make an error of fact or law is never misconduct: Moran v Lloyds [3] QB 542;
(b) that the most common form of misconduct is to decide a case on a basis that was not pleaded or argued or in issue: eg Interbulk Ltd v Aiden Shipping Co Ltd, “The Vimeira” [4] 2 Lloyd’s Rep 66; Modern Engineering (Bristol) Ltd v C Miskin & Son Ltd [1981] 1 Lloyd’s Rep 135; Fox v P G Welfair Ltd [1981] 2 Lloyd’s Rep 514;
(c) that the way in which costs are dealt with also tends to give rise to such applications; eg where the arbitrator gives no reasons for apparently not giving effect to a sealed offer as in The Ios [7] 1 Lloyd’s Rep 321.
A question which has yet to be fully elucidated is whether the distinction between misconduct and procedural mishap has any practical consequences. The distinction itself is finely illustrated by Harrison v Thompson [9] 1 WLR 1325. The case involved an arbitration as to the value of a shareholding. Both parties made Calderbank offers. Both expected to have the opportunity to address the arbitrator on the question of costs once the award had been published but omitted to ask him to make an interim award. He ordered that each party bear its own costs. There was clearly no misconduct here. Nevertheless Knox J ordered that the award be remitted for reconsideration of the question of the costs on the ground of procedural mishap. He held that the test for remission is “risk of injustice”, injustice being defined as a situation where a procedural mishap has caused relevant material not to be put before the arbitrator; deliberate choices, however, cannot constitute procedural mishap. The distinction may affect the onus of proof. It seems that if misconduct is shown, the onus is upon the party claiming that it did not make any difference to the outcome to establish that, whereas, in a case of mere procedural mishap, there may be an onus on the party seeking to have the award remitted to establish that the mishap did make a difference: see the Aros [1978] 1 Lloyd’s Rep 456. In practice, however, any incident giving rise to a reasonable likelihood of injustice will probably result in the award’s being remitted.
Inadvertence
One should also mention two ways of making an award judge proof inadvertently. The first is to fail to ask for reasons at the appropriate time. The other is to fail to appeal/apply to set aside or remit in time.
If the arbitrator does not give reasons the court has power to order him to do so or to do so more fully, but can exercise this power only where, before the award was made, one of the parties gave notice that a reasoned award would be required or where there is some special reason why such a notice was not given. An oversight or misunderstanding by one of the professional advisers of the would-be appellant is not a special reason: Leeds Permanent Building Society v Latchmere Properties Ltd [9] 1 EGLR 140; [1989] 20 EG 128.
Appeals and applications to set aside/remit have to be brought by originating motion within 21 days of the publication of the award: RSC Ord 73 rr 2 and 6. The courts are extremely strict about this time-limit and will extend time only when:
(i) good cause is shown;
(ii) there is no possibility of prejudice; and
(iii) the delay is minor
(per Bingham J in Zermalt Holdings SA v Nu-Life Upholstery Repairs Ltd [5] 2 EGLR 14). An award is published when the arbitrator tells the parties it is ready: Bulk Transport Corp v Sissy Steamship Co “The Archipelagos” [1979] 2 Lloyd’s Rep 289, said to apply to rent review arbitrations in South Tottenham Land Securities v Millett (1983) 268 EG 703 at p 706. So delay in paying the fee and taking up the award may sacrifice the right to appeal/apply. Failure to warn the client of the shortness of this time-limit is actionable negligence: Corfield v Bosher & Co [1990] NPC 1.
Conclusion
My conclusion is that, in practice, an error of procedure is more likely to lead to an award’s being impugned than an error of law. This conclusion, based on a survey of the cases, accords with my own experience that
(a) arbitrators not infrequently have to be guided back on to the straight and narrow procedural way in the course of an arbitration, but
(b) anticipated questions of law often evaporate upon a consideration of the valuation implications or, when they do not, are usually very adequately dealt with by the appointed arbitrators.
(1) Published at (1992) 263 EG 781 and 867.
(2) Celebrated because the cases in question were Estates Projects Ltd v Greenwich London Borough Council (1979) 251 EG 851, F R Evans (Leeds) Ltd v English Electric Co Ltd (1977) 245 EG 657, New Zealand Government Property Corp v HM&S Ltd [1] 1 WLR 870, Pivot Properties Ltd v Secretary of State for the Environment (1979) 256 EG 1176, and Plinth Property Investments Ltd v Matt, Hay & Anderson (1978) 249 EG 1167.
