The Settlement Saddlepoint: Why Employment Claims Are Hardest to Settle When It Matters Most

In game theory, the saddlepoint is the equilibrium where settlement becomes rational. In employment litigation, reaching it requires information that can only be acquired by spending the money the settlement was supposed to save.

Drag to rotate. The surface curves up along the information axis (more information, better settlement prospects) and down along the costs axis (more expenditure, less surplus to divide). The saddlepoint sits at the unstable equilibrium between the two. The red line traces the typical case path: costs accelerate faster than information, pulling the case off the ridge and into the trough.

ACAS early conciliation opens before either side has seen a single document. The claimant has a grievance outcome and a solicitor's letter. The respondent has a file note from HR and a vague recollection of what was said in a meeting three months ago. Both are invited to settle.

This is, on any rational analysis, absurd. Neither party knows what the other holds. Neither can quantify the claim with any confidence. The claimant does not know whether the respondent's internal emails support or destroy the narrative. The respondent does not know whether the claimant kept a contemporaneous diary, recorded conversations, or has a medical report connecting the conduct to a psychiatric injury. They are being asked to price a claim in the dark.

Yet this is also, on any rational analysis, the cheapest moment to settle. Legal costs are minimal. Management time has not been consumed. The emotional investment is still recoverable. The total surplus available to divide between the parties, the pot of money that would otherwise be burned on litigation, is at its maximum.

This is the settlement saddlepoint paradox. And it is the defining structural problem of employment litigation.

What the Saddlepoint Is

In game theory, a saddlepoint is the equilibrium in a strategic interaction where neither party can improve their position by unilaterally changing strategy. Applied to litigation, the saddlepoint is the moment at which both sides' expectations of the outcome converge sufficiently to permit a negotiated resolution. When the claimant's minimum acceptable settlement overlaps with the respondent's maximum willingness to pay, a "zone of possible agreement" exists. Settlement becomes rational.

The problem is that convergence requires information. Each party enters litigation with their own subjective estimate of success. The claimant, naturally, believes their case is strong. The respondent, equally naturally, believes it is weak. This "mutual optimism" keeps the parties apart. It is only through the forced exchange of information, through disclosure, witness statements, and the stress-testing of positions at preliminary hearings, that subjective beliefs are updated and expectations begin to align.

The paradox is circular. To settle rationally, the parties need information. To get information, they must litigate. To litigate, they must spend the money that settlement was supposed to save.

The ET Timeline as a Surplus-Destruction Machine

Consider the procedural stages of a typical employment tribunal claim mapped against two variables: the information gap between the parties (how far apart their assessments of the case are) and the settlement surplus (the total costs both sides would save by settling rather than proceeding to trial).

At the ACAS early conciliation stage, the surplus is at its peak but the information gap is vast. By the time witness statements have been exchanged and the parties finally know enough to settle rationally, 80% or more of the available surplus has been consumed in legal fees, management time, and the opportunity cost of distraction. The saddlepoint, the zone where settlement becomes economically rational, sits in the most expensive part of the litigation timeline.

Mutual Optimism and the Fog of Litigation

The engine driving this paradox is what economists call "mutual optimism." Both parties overestimate their chances. The claimant remembers the meeting vividly and has a sympathetic narrative. The respondent has a paper trail of performance management and believes the tribunal will see through the claim. Each side, working from incomplete information, constructs a version of events in which they win.

This is not irrational. It is a predictable consequence of information asymmetry. The claimant knows things the respondent does not (the private impact on health, the conversations with colleagues, the text messages). The respondent knows things the claimant does not (the internal deliberations, the comparator data, the advice from HR). Each party's confidence is calibrated to their own information set, not to the complete picture.

Every procedural step in the tribunal process is designed, whether intentionally or not, to erode mutual optimism. The ET3 response reveals the respondent's case theory for the first time. A preliminary hearing on strike-out or deposit order (Employment Tribunal Rules of Procedure 2024, Rule 39) forces an early judicial assessment of prospects. Disclosure produces the documents each side feared. Witness statements reveal how the key witnesses will actually present. Each stage narrows the information gap and forces the parties to update their beliefs.

The Procedural Pressure Points

The UK employment tribunal system builds in several mechanisms that attempt to force settlement before the surplus is exhausted. Each operates at a different point on the timeline, and each has structural limitations.

ACAS Early Conciliation. The statutory gateway. It operates before any procedural costs have been incurred but also before any information has been exchanged. It works best where the facts are broadly agreed and the dispute is about valuation: how much is this worth? It works worst where the parties disagree about what happened, because neither side has the information needed to assess credibility or documentary support. The paradox is at its sharpest here.

Section 111A and protected conversations. Pre-termination negotiations under s.111A of the Employment Rights Act 1996, as interpreted in Faithorn Farrell Timms LLP v Bailey [2016] IRLR 839, allow employers to make settlement offers before any dispute has crystallised. The inadmissibility protection extends to internal discussions about the offer. This is a mechanism designed to locate the saddlepoint before litigation begins, but it depends on the employer making a realistic offer at a point when neither side has tested the evidence.

Deposit orders. Rule 39 allows a tribunal to require a party to pay a deposit of up to £1,000 as a condition of pursuing an allegation with "little reasonable prospect of success." This is the tribunal system's attempt at a separating equilibrium: filtering weak claims from strong ones before the expensive disclosure and witness statement stages. If the party loses on the deposited point, they face a presumption of unreasonable conduct for costs purposes. It is a blunt instrument, but it operates at the right point in the timeline, between the preliminary hearing and full disclosure.

"Without prejudice save as to costs" offers. In Kopel v Safeway Stores plc [2003] IRLR 753, the EAT confirmed that while the Calderbank principle (automatic costs consequences for rejecting a reasonable offer) does not apply in the employment tribunal, a party's rejection of a settlement offer is evidence of unreasonable conduct that a tribunal may consider when deciding whether to award costs. This is weaker than the civil courts' Part 36 regime, which imposes automatic cost consequences. The limited costs-shifting environment of the ET means the financial penalty for failing to settle at the right moment is muted.

Why the ET Costs Regime Makes It Worse

In the civil courts, costs follow the event. Lose, and you pay the winner's costs. This concentrates minds. It makes the failure to settle at the saddlepoint financially dangerous, because the costs incurred after a reasonable offer are likely to be shifted onto the party who rejected it.

The employment tribunal operates differently. Costs are exceptional. Rule 74 requires a finding of unreasonable conduct, and even then the award is discretionary. As Mummery LJ put it in Yerrakalva v Barnsley MBC [2012] IRLR 78, the "vital question" is whether the party's unreasonable conduct caused the costs in question to be incurred. The tribunal must look at the "whole picture," not punish parties for single misjudgments.

This is humane. Employment disputes are not commercial transactions between sophisticated parties of equal means. Many claimants are litigants in person. Many respondents are small employers. The costs regime reflects the policy that access to justice should not be deterred by the threat of a ruinous costs order.

But it also means there is very little financial incentive to settle early. The claimant who rejects a reasonable offer at the ACAS stage and ultimately recovers less at trial faces, at most, a discretionary costs order of the kind discussed in Vaughan v London Borough of Lewisham [2013] IRLR 716. The respondent who refuses to engage with early conciliation and forces the claimant through twelve months of litigation may face no costs consequence at all if the claim succeeds on a narrower basis than pleaded. The saddlepoint paradox is not just an information problem; it is reinforced by a costs regime that does not penalise the failure to find it.

Implications for Practitioners

None of this is fixable. The paradox is structural. But awareness of it changes how you advise.

1. Price the unknown, not the known. At the ACAS stage, the temptation is to wait for more information before settling. Resist it. The information you are waiting for will cost more to acquire than it is worth. Make a realistic assessment based on what you know and factor in a discount for uncertainty, rather than burning the surplus to reduce the uncertainty by 20%.

2. Treat every procedural step as a settlement event. Each stage, the ET3, the preliminary hearing, the disclosure list, the witness statements, changes the information landscape. Each is an opportunity to reassess. Practitioners who only think about settlement at formal mediations or ACAS conciliation miss the incremental convergence that happens at every procedural turn.

3. Use deposit orders strategically. For respondents, a deposit order application at the preliminary hearing stage is the most efficient tool in the system for breaking the paradox. It forces an early judicial assessment, puts financial pressure on weak points, and creates a costs hook for later. For claimants, surviving a deposit order application without a deposit being imposed is a powerful signal that strengthens the negotiating position.

4. Make offers early and make them real. A "without prejudice save as to costs" offer made at the ACAS stage, even if rejected, creates a marker for any later costs application. Under Kopel, it is not automatic, but it is evidence. An early, realistic offer also forces the other side to engage with valuation before their mutual optimism has fully calcified.

5. Accept that you are settling in the dark. The saddlepoint paradox means that by definition, the optimal moment to settle is before you have the information you want. Every settlement is a bet. The question is not whether you are betting, but whether you are betting with the surplus still on the table.