Thanks to the Woolf reforms and a recent chancery ruling, parties that win an arbitration on one good point but bring up other issues that have no legal merit may find themselves paying the costs of the loser.
Mr Justice Neuberger, a judge of the Chancery Division, has come up with a decision about costs that seems to be an absolute model of reasonableness, but may create some serious problems in construction arbitration.

Antonelli vs Allen had nothing at all to do with the building industry, but the end result sounds familiar. The plaintiff won, but not on all issues. The judge decided that he should consider whether Antonelli had been reasonable in bringing the points that he had lost, the time that had been taken up on those issues, the increase in paperwork and the like. The upshot was that he decided the successful plaintiff should pay 75% of the defendant's costs.

That all sounds perfectly reasonable, doesn't it? After all, it would be unfair to make a defendant pay all the costs of a huge action with lots of different points, most of them bad ones, just because the claimant managed to get home on one good point. With the overriding objective of the Civil Procedure Rules in mind, parties should be discouraged from running up huge bills pursuing issues without merit.

The same approach should be appropriate in arbitration. If one party brings a complex claim and succeeds only on a small part of it, the arbitrator may feel that it would be quite wrong to let costs "follow the event". The bill run up on the bad points might dwarf the costs on the good ones. The respondent may have won hands down on several aspects of the case, but because the claimant has a net win overall the "event" is in its favour and the traditional approach would mean that it would recover the costs of the bad points as well.

Arbitrators have always felt able to avoid the worst extremes of this sort of injustice, awarding the net winner a percentage of costs or ring-fencing some issues and making no order for costs on them. The Woolf approach to costs encourages them to take this rather further than they might have done a few years ago.

The typical main contractor vs employer dispute involves a long list of issues. There will be several disputed variations and a catalogue of events that are said to entitle the contractor to extensions of time and payment of loss and expense. There may be a counterclaim relating to liquidated damages or defects. Some of the claims have merit and some have none. If the arbitrator really wants to achieve perfect justice, he may be tempted to deal with the costs of each issue or group of issues separately. He might feel tempted to award the contractor the costs of all the issues in which it succeeded, and make the contractor pay the employer's costs of the issues in which it failed.

  • In a recent Chancery case, the winner paid 75% of the loser’s costs because several of its points had no legal merit
  • Arbitrators could do the same in building disputes

  • There was a discussion of just this sort of point at a conference of construction arbitrators in Llandudno last October. A hypothetical case was constructed, with one party being successful in some heads of claim but failing in others. It was not a particularly complicated case as construction arbitrations go, but at least a dozen different answers were offered by the audience of highly experienced and respected arbitrators. They ranged from "winner takes all" to "a plague on both your houses" with no order for costs at all. There were sophisticated part costs orders and there was a whole range of percentages knocked off and added on.

    What was absolutely certain was that there was absolutely no certainty about what an arbitrator would do in such circumstances. This makes life very difficult for those of us who try to make our way in the world by advising either contractors or employers on how to conduct disputes, and how much arbitration is likely to cost.

    It also raises the very serious question of how a respondent in an arbitration should protect itself. The traditional advice was to write a letter "without prejudice save as to costs" offering to pay what the respondent thought the claim was worth. If the claimant succeeded in persuading the arbitrator to award a sum greater than the offer, it could expect to recover costs. If it failed and the arbitrator awarded less than the offer, it would end up paying the respondent's costs from the date of the offer. This sort of letter was mysteriously known as a "Calderbank", named after an obscure case in the divorce courts for reasons that would take too long to explain here.

    But if the arbitrator is going to start awarding the costs of one issue one way, the costs of another issue another way, and perhaps making no order for costs at all on other issues, these Calderbank letters will have to become far more sophisticated, with separate offers being made in respect of some, or possibly even all, the issues in the case. As the investigations into each item develop, the views on merits and values may change considerably, and a host of revised specific Calderbank letters will appear. The arbitrator's task in dealing fairly with the costs issues will become even more complicated than the issues in the arbitration proper.