Two neutrals, one week, two comparable disputes. In the arbitration, the first procedural conference runs the better part of two hours. The timetable is argued over. Memorials or pleadings. Whether there will be document production and on what model. A Redfern schedule and who populates which column. Expert conferencing, hearing length, transcript, site visit. At the end of it, a written instrument governs the reference, and every subsequent step is measured against it. In the mediation, the pre-mediation call lasts twenty minutes. Rooms and venue. Who is attending and whether the clients will be there in person. Position papers by the Wednesday, 10 pages, exchanged or confidential. Whether there will be an opening session, a question usually resolved in the negative. Then the date is confirmed.
One neutral designed a process. The other confirmed a booking.
Last month in this column I argued that the one-day model has quietly stopped being a format and become a definition, and that definition is now colliding with the disputes that international mediation is asked to resolve. This article addresses a significant consequence I did not have room to develop. The day has not merely constrained what mediators do but has also shaped what mediators know. A profession that performs the same process every time never has to learn how to design one.
The Instrument We Never Wrote
Arbitration’s focus on process design is not innate, and arbitrators are not cleverer than mediators. The competence exists because arbitration produces an artifact. Procedural Order No. 1 is drafted, negotiated, circulated, imitated, criticized, taught, and benchmarked. The IBA Rules and the Prague Rules are effectively a public debate about process design, conducted in print, over decades, in front of the entire field. Because arbitration writes its procedure down, arbitration can argue about it; because it argues about it, it can improve. The apprentice arbitrator learns by reading a hundred procedural orders before drafting one.
Mediation has no such instrument. What we design, we design in our heads, often on the way to the venue, and we tell nobody. There is no artifact, therefore no record. No record, therefore, no critique; no critique, therefore no transmission. Whatever design intelligence the field’s best mediators possess, and it is considerable, largely retires when they do. Confidentiality is usually blamed for this, and it is a poor excuse. The architecture of a process is not the parties’ commercial information. The shape of a mediation can be described in full without disclosing a word of what was said inside it.
How the Day Trains It Out of Us
The deskilling is structural rather than personal, and it operates through at least five mechanisms, none of which is anybody’s fault.
Training and accreditation assess the performance of a day. The assessed role play is a day in miniature: the opening, the caucus, the reality test, the corridor, the close. The competences examined are room competences, and they are examined superbly. No mediator I know of has ever been assessed on a process design. We certify performers and then wonder why we have performers.
Current fee models price attendance rather than thought. The daily rate pays for a mediator’s presence in a building. Diagnostic work, stakeholder mapping, sequencing and the drafting of interim instruments happen, if at all, in unbilled hours. Unbilled work becomes a courtesy, and courtesies are done thinly and late. A fee model is a curriculum: it tells practitioners which parts of the job the market believes are the job.
Selection reinforces both. Parties choose mediators on settlement statistics, sector familiarity, and a reputation for closing. Nobody, in my experience, has been appointed on the strength of a process proposal, because nobody is asked for one. A market that does not request design does not receive it and cannot then complain of its absence.
Feedback is measured at the wrong point. The only outcome the field records is settlement at the door. A mediator who mismanages the architecture of a complex case but closes at nine in the evening is recorded as a success; one whose phased process settles four weeks later at a scheduled review meeting may not be recorded at all. We are optimising against a metric structurally incapable of seeing the competence I am arguing we should build.
Apprenticeship depends on observability. A day can be watched. You can put a pupil in the corner of a room and let them see a professional work, and much of the craft transmits without anyone naming it. Design happens over weeks, mostly in silence, in decisions about who must speak to whom and in what order and what must be known before anyone meets at all. Tacit knowledge passes by being visible. Design is invisible.
The cumulative effect is a profession fluent in choreography and largely devoid of architecture. That is not a criticism of mediators. It is a description of the incentives we have built around them.
The Design Tradition Next Door
What makes this omission strange is that the discipline already exists. Commercial mediation simply never went to fetch it. Public policy, environmental, and community mediation have practised process design as a named competence for four decades. The convening or conflict assessment phase developed in that world is precisely what commercial practice lacks: a period before any joint session in which the neutral interviews the stakeholders, maps interests and constituencies, identifies who must be at the table and who need only be consulted, tests whether the matter is ripe, and produces a written recommendation on whether and how to proceed at all. Professor Lawrence Susskind and his colleagues built handbooks around it. Whole practices are organized on it.
The scholarly scaffolding is older still and is not obscure. Frank Sander’s multi-door courthouse was a design proposition before it was an institutional one: match the dispute to the process. Sander and Goldberg’s fitting the forum to the fuss is, quite literally, a design manual. The dispute systems design tradition of Ury, Brett and Goldberg taught organizations to think in sequences of processes, and in the movement between interests, rights and power, rather than in the selection of a single procedure.
Commercial mediation borrowed the neutral from this literature and left the design behind. We took the third party, the confidentiality and the caucus, and we left the assessment phase in the box. We did so at precisely the moment when the disputes arriving at our door, which are multi-party, cross-border, politically constrained, and technically dense, have come to resemble the disputes the design tradition was built for.
What Design Actually Requires
It is worth being concrete, because process design is a phrase that invites abstraction. Designing a mediation means answering, before the first joint session, a series of questions the single-day mediation never has occasion to ask.
The first concerns ripeness. What does each side need to know before it can responsibly price its own risk, what would supply that knowledge, who has that knowledge, how long would supplying it take, and who supervises the exchange? A day offers no mechanism for this; a designed process treats it as the opening phase.
The second concerns authority. Who can actually say yes, on what cycle do they meet, what papers do they need in front of them, and how far in advance? This deserves a column of its own.
The third concern is geometry. In a multi-party case, which conversations must precede which others, what can be handled bilaterally, what must be plenary, and where does a partial settlement unlock the matter rather than stranding a party who then has every incentive to obstruct it?
The fourth concerns sequence, and what gets signed along the way. Phases with end dates, standstills, preservation measures, partial and interim agreements, frameworks initialled subject to confirmation, scheduled review points. These are the tools that make time a design variable instead of a constraint inherited from a diary.
The fifth concerns the endgame. What must this settlement survive? Documentation, board or shareholder approval, regulatory or sanctions clearance, tax treatment, and where there is a continuing relationship, the governance framework that will carry it for the next five years. Competent design works backwards from all those issues, rather than discovering them at eleven at night or not at all.
Behind all five sits the most uncomfortable question of all, which is whether mediation, alone, is the right process. Perhaps a valuation gap needs an expert determination whose outcome feeds into a negotiation. Perhaps an early neutral evaluation on one contested point of law would release everything downstream. Perhaps the case needs a mediation now and an arbitration on two carved-out issues later. A neutral trained only to run a day, and paid only when a day is booked, is not ideally placed to be a part of that discussion. Arbitration surrounds its neutrals with elaborate machinery for identifying conflicts of interest. Mediation has built nothing at all around the recommendation of process, because we have never quite regarded process as a choice.
A Modest Instrument
The remedy is not a rulebook. It is a page. There is no reason a mediator, after a proper convening phase, should not produce a short, written process memorandum, setting out the diagnosis, the phases and their dates, the information steps and who supervises them, the participants and the approval architecture behind them, the interim instruments contemplated, and the implementation plan. One to three pages, agreed with the parties, revisable by consent. The field will resist calling it Procedural Order No. 1, and it should. The name matters far less than the fact of writing it down.
The virtues of such a process and document are mostly second order, which is exactly the point. A written design can be discussed with the parties before it is applied to them, which improves it and, incidentally, transfers ownership of the process to the people whose dispute it is. It can be revised deliberately rather than abandoned quietly. It can be shown, anonymized, to a colleague or a training faculty. It can be criticized. Fees can be structured against phases rather than days, so that design becomes paid work and therefore real work. And within a decade the field would possess what arbitration has had for thirty years: a visible body of practice about which serious professionals can disagree in public.
There are honest objections. Over-design is a genuine vice, and a mediation buried in architecture is as useless as an action buried in interlocutory applications; every practitioner has met a process that became a substitute for a decision. Front-loading costs money that some parties do not have. Many disputes need nothing more than a well-run day, and convening a design phase for a two-party money claim with both principals available and both insurers reserved would be self-indulgent. All of that is true. But the argument is not that every dispute requires elaborate design. It is that every dispute deserves a conscious decision about its design, and that a profession which never makes such decisions deliberately will not make them well on the occasions when it must. The objection to designing every dispute is sound. The objection to designing none is unanswerable.
It is also said that the best mediators design by instinct and would only be slowed by writing it down. Having watched a good number of them, I accept the premise and reject the conclusion. Musicians improvise because they have studied form, not instead of studying it. What the field lacks is not talent. It is theory, and the shared vocabulary that theory produces, for what we are actually doing when we decide that these two people should meet on Thursday and those four should not meet at all until the delay analysis has landed.
The Skill After the Day
The previous column argued that mediation should stop being an event and become a process. The harder point is that a profession cannot competently manage a process it has never been trained to design. The single day gave us superb performers: mediators who can read a room, hold a silence, absorb an insult without returning it, and carry a derisory offer up a corridor with their expression under perfect control. Those are real skills, and they will continue to matter. But the disputes now arriving, and the courts, institutions, and governments now sending them, are putting a different question to the neutral. Not whether you can close this today, but what process this dispute actually needs and whether you are capable of building it.
That is an architect’s question. We have spent forty years training dancers and training them very well. The competence the field must now add will not be learned in the room. It will be learned before anyone enters the room.