The costs system continues to crumble into incoherence and injustice. Proportionality has been part of the CPR since 2013, but has never been defined beyond the more or less meaningless words of CPR 44.3(5), which effectively says costs are proportionate if they are proportionate. In his preface to the 2013 edition of the White Book Jackson LJ said that guidance would be needed from the Court of Appeal. As the 2016 edition shows, that guidance is still awaited. In the meantime litigants are being knocked around the ring by random decisions at first instance, though at least one party in three cases decided since last November is seeking leave to appeal.
In Hobbs v Guy’s and St Thomas’ NHS Foundation Trust
 EWHC B20 (Costs), decided in November 2015, Master O’Hare had to consider a bill of £32,000-odd in a clinical negligence case that had settled for £3,500. Applying the traditional tests, Master O’Hare assessed the costs reasonably incurred at around £11,000. He then turned to consider proportionality.
In February 2015 Leggatt J had given some guidance on the meaning of proportionality in Kazakhstan Kagazy PLC v Zhunus
 EWHC 404 (Comm) : ‘The touchstone is not the amount of costs which it was in a party’s best interests to incur but the lowest amount which it could reasonably have been expected to spend in order to have its case conducted and presented proficiently, having regard to all the relevant circumstances.’ As Master O’Hare remarked at , he had already decided that £11,000-odd was ‘the lowest amount which the claimant could reasonably have been expected to spend in order to have this case conducted and presented proficiently, having regard to all the relevant circumstances’ so strict application of Leggatt J’s test would make reasonable costs automatically proportionate. He accordingly distinguished Kazakhstan Kagazy
. The sums at stake in Kazakhstan Kagazy
had been so enormous that no costs, however high, could ever amount to a significant proportion; in case like Hobbs
, on the other hand, where the costs had – quite properly – exceeded the recovery, different considerations must apply.
Concluding that the costs were indeed disproportionate, Master O’Hare faced a choice: he could cut an arbitrary proportion of the reasonable costs, or he could disallow specific items. Declining to make a random cut, he instead used hindsight to disallow items which, though reasonable at the time they were incurred, had been shown to be disproportionate by the final outcome of the case; he accordingly reduced the bill by a further £1,200. Hindsight breaches, of course, one of the most basic principles of traditional costs assessment, but Master O’Hare held that the old rule in Francis v Francis and Dickerson
 3 All ER 836 has been overtaken by proportionality.
Although it left the claimant with £2,300, this approach certainly has the potential to work injustice. Put crudely, a clinical mishap may cause £50,000-worth of damage, but finding out that only a tenth of that damage can be attributed to negligence often takes a long time, involves several experts and costs a great deal of money. In this area of work in particular the claimant with a poor claim is acting no less reasonably, and litigating no more frivolously, than a claimant with a cast-iron case: on any account even the unmeritorious claimant has suffered genuine damage while in the defendant’s hands, but without expensive professional advice from both doctors and lawyers has no means whatever of judging whether he is the victim of negligence or mischance. Where the cost of that advice should fall may be an awkward question, but it is going to fall increasingly on a patient who is damaged, who has not contributed to his loss in any way and who has conducted his claim entirely reasonably.
looks unjust, two subsequent cases look even worse. In BNM v MGN Ltd
 EWHC B13 (Costs), decided at the beginning of June, a primary school teacher took action against a newspaper publisher to prevent the use of data taken from her mobile phone; the phone had gone astray accidentally, but had found its way into the newspaper’s hands because it revealed a long-standing relationship with a Premiership footballer. After settlement a detailed assessment found the reasonable base costs to be £62,000-odd, a figure that additional liabilities took to £167,000-odd.
When he came to apply the proportionality test, Master Gordon-Saker held that it must cover not only base costs, but also additional liabilities such as the ATE premium. The figures finally allowed were roughly one half of the assessed reasonable costs.
Although he went through the statutory criteria, it is impossible to see how the Master navigated from the reasonable costs to the proportionate costs. Why one half? Why not a third? Or two-thirds? ‘In these circumstances base profit costs of £46,000 and base counsel’s fees of £14,000 must be disproportionate under the new test, being over three times the amount of agreed damages, and covering work which fell far short of trial. In my judgment costs of about one half of those figures would be proportionate.’ 
Halving the ATE premium is particularly hard to explain. At  the Master had expressly found ATE insurance to be necessary and the premium reasonable at £58,000; condemning as disproportionate a necessary item obtained at reasonable expense is effectively the same as saying the Court should not have been troubled with this litigation in the first place, since on that premise the case could never have been profitable.
May & anor v Wavell Group Plc & anor
 EWHC B16 (Costs) is probably an indication of the direction in which the courts intend to take proportionality. Describing proportionality as a ‘blunt instrument’ at , Master Rowley characterised Leggatt J.’s test as ‘still too generous to the receiving party’ where the sums in issue are modest. ‘In many modest cases’ the sum to be paid in costs will not be what is ‘required to achieve justice in the eyes of the receiving party but only a contribution to that receiving party’s costs.’
was a claim in nuisance. The claimants accepted £25,000. Their reasonable costs were assessed at a shade under £100,000. The award was £35,000. Again, there is no way of stating why the proportionate sum was £35,000 rather than £25,000 or £50,000. Dr and Mrs May brought a claim, achieved a settlement, but were left with £40,000 less than they had started with.
The unspoken but underlying assumption that parties should use ADR rather than court proceedings for modest claims will not withstand a moment’s thought. Mediation or any other form of negotiation works only if there is an alternative worse than the proposed agreement. If everyone knows from the outset that the claimant will not sue because a loss is certain whatever the outcome, there is no adverse alternative, so why would any sane defendant negotiate?
The courts can, and increasingly do, force parties to the negotiating table through penalties for failure to engage in ADR. What they cannot do, however, is force parties to negotiate once at the table. If a potential defendant simply sits through an ADR session, saying in effect ‘I have nothing to offer, because proportionality prevents you from issuing proceedings’, the courts cannot penalise that behaviour. To brand that behaviour as unreasonable would force the courts not only to determine the merits of the original claim, but also to identify the party responsible for the failure of negotiations.
Common sense says that will never happen. Such an investigation would destroy the ability to negotiate without prejudice and would turn on disputed conversations of which there are no records whatever.
The message is clear. Do not sue. Forget mediation. Unless your loss runs into at least six figures, and probably much more, swallow your pride and put it down to experience.
In the short term a claimant’s best weapon is an early, realistic offer under Part 36. If the claimant beats his own offer at trial, CPR 36.17(4)(b) gives him in most circumstances indemnity costs from the date on which the offer expired. Indemnity costs, of course, make proportionality irrelevant: the difference between a winning Part 36 offer and an also-ran can therefore easily amount to many tens of thousands of pounds.
At one level this provision goes a short distance towards mitigating the injustice caused by BNM
and to that extent modifies the message coming from the courts. If you have a modest claim, the modified message now runs, make a realistic Part 36 offer before you do anything else. That is the only way to give the defendant an incentive to negotiate. You will almost certainly recover only a modest proportion of your modest claim – because your threat to issue proceedings will be hollow unless beating your own offer is a racing certainty – but something is better than nothing and overall the transaction will show a profit.
In the longer term, however, some form of fixed-costs regime on the German model seems likely to replace the current lottery: Jackson LJ has proposed fixed costs in cases up to £250,000, which he describes as the foothills of multi-track litigation. Unless they come with a really drastic procedural and evidential simplification of such cases, however, fixed costs will do no more than replace the random losses of BNM
with fixed losses. The upshot will be the same: do not sue, and do not mediate, modest cases because you will never come away with money in your hand.
If procedure and evidence were open to drastic simplification, one might think the Woolf and Jackson reforms would have achieved it. For the smallest claims the proposed online court may offer radical change and turn out to be not very different from the original rough-and-ready County Court of 1846. That will still leave a killing-zone of dead ground between £25,000 and the foothills of multi-track litigation at £250,000.
As I have said before, watch this space!