Posted on: 2 October 2026
I once followed a merger that the lawyers on both sides had framed from the first meeting as one more thing to be mediated. The unions were pushing to block it and between the two companies sat the usual gap between what was being asked and what was being offered, which the lawyers were getting ready to narrow over a few months of drafts until somebody grew tired enough to sign. It went differently because we built the negotiation around what each party actually needed, unions included, and left the figures in the opening letters to one side. The frame shifted when a variable appeared on the table that none of those letters contained, namely what would happen to the people the merger would leave out. Anyone who lost their job would be placed with supplier companies of the business being merged, in the same area, after proper retraining. At that point the deal the unions had wanted to stop became a yes for them too. Everyone at the table came out better off and the lawyers on both sides found themselves almost in the shadows, checking the compliance of an agreement that had been built somewhere else. They did their job very well. It just happened to be a different job, and that is where I want to start.
Lawyers brought in to settle a dispute or to close a transaction nearly always try for an agreement and call the process negotiation, but it almost never is. It is bargaining, even though the two words keep ending up in the same sentence without anyone noticing the difference.
Bargaining works on a single variable. I ask for a hundred and twenty, you offer forty and we stop at the point where carrying on costs more than the irritation of closing. Whatever one side gains the other loses and at the end the table holds exactly the value it held at the start, minus the fees. Walton and McKersie, studying American union contracts in 1965, called this distributive bargaining and set it beside the integrative kind, which starts from an almost banal observation: the parties never put the same weight on the same things. One needs cash now, the other needs time or needs the whole affair kept quiet. When there is more than one variable and each side weighs them differently they can be traded, until the final total ends up larger than the one you started with.
Put like that it sounds like a textbook technique, when in fact it is mostly work. To negotiate integratively you need to know a great deal about the other side, what it really needs and how much it is prepared to spend to get it, and you need to know it before you sit down. That information serves two purposes pulling in opposite directions, because the same knowledge that lets you build an exchange good for everyone also tells you how far the other party can be pushed. A good negotiator handles that tension with precision, revealing what helps the table grow and keeping back what would let the other side walk off with the larger share. Meanwhile he works on the alternative, the well known BATNA, because strength at the table depends on how easily you can afford to stand up and leave. A credible alternative cannot be improvised and sometimes takes weeks of proper intelligence work on a company or on a person. All of this costs time, which is precisely the resource a lawyer's professional model tends not to spend.
The easy explanation is that lawyers want to do less work, and it is only partly true. If that were the whole story, a lawyer billing by the hour would have every reason to drag things out and explore elaborate solutions, yet he too nearly always ends up trading numbers. The deeper reason is an asymmetry of risk. If a lawyer proposes a creative solution and it goes wrong, the responsibility is his and everyone can see it. If he sticks to legal compliance and a numerical compromise, nobody will ever ask him about the value that was never created, because that value appears in no document. The damage is visible and the missed gain is invisible. It is skin in the game turned inside out, where the skin is there all right but sits entirely on the side of caution.
In company sales the mechanism can be seen almost under a microscope. The buyer's lawyer answers for the representations and warranties and for the indemnity thresholds. He does not answer for the price and still less for what the two parties might have built together, such as a seller who stays on for two years with a variable share tied to results, or a supply contract that is worth more to the buyer than the discount he is trying to squeeze out. Solutions like these only appear if someone goes looking for them, and the lawyer, by mandate and by incentive, goes looking for risk. He does it well, often very well, but anyone expecting him to create value as well is asking the goalkeeper to score.
So the gap is left by the client. Plenty of business owners hand a deal or a dispute to their lawyers the way you hand a car to a mechanic, and they get back exactly what a mechanic knows how to deliver, which is a roadworthy vehicle and a bill. Pushing the lawyer to be creative achieves little, because you are asking him to take on a risk nobody pays him for and his training has taught him to avoid, so what comes back is usually a cautious proposal dressed up as an idea. What is needed is to fill the empty seat. The client keeps hold of the negotiation, or puts someone paid on results in charge of it, and leaves the lawyer to do what he does better than anyone: guard the litigation alternative and make watertight whatever is finally signed.
In digital cinema the conversion of European and American screens was for years a negotiation stuck on a single question, which was who pays for the projectors. Exhibitors did not want to fund an investment whose savings went mostly to the distributors, and distributors did not want to pay for equipment they would not own. Every meeting ended where it had started. The deadlock broke when someone looked at both sides' accounts rather than at their demands. The studios were spending more than a thousand dollars on every film print, so they could afford to pay a fee for each title booked in digital and with that fee the exhibitor paid off the equipment. It was called the virtual print fee. It was a profit and loss idea, and no letter before action would ever have produced it.
The weight of legal training is felt even when nobody is going to court. A judge can award a sum or declare a right, so anyone trained to turn every conflict into a claim to put before a judge ends up reasoning inside that menu outside the courtroom too. Settlement becomes a judgment delivered early, discounted for time and risk and built on the same few items. Italy has even given the habit an official name. Since 2014 the procedure that obliges parties to try for an agreement before certain claims can proceed has been called assisted negotiation, and in practice it consists largely of an exchange of letters and figures.
Then there is a case where the client cannot take the seat even if he wants to. Some time ago I received a fine in Switzerland that I considered unfair, and since I had been paying for legal expenses insurance for years I opened a claim, with the satisfaction of someone who has finally found a use for a premium paid into the void. The insurer did what nearly all of them do in these cases, which was to send a letter. One letter, polite, with the legal references in the right places, that I could have written myself in half an hour at the kitchen table. Then the file was closed and the fine stayed where it was. To be fair, there is nothing to trade with an authority that has fined you, so integrative negotiation was never on the cards, but what struck me was something else, namely who was deciding how much work my case deserved.
With legal expenses insurance the legal work is paid for by the insurer, which collects a fixed premium and makes money when each file costs as little as possible. The real principal of the in-house lawyer handling the claim is therefore whoever pays his salary, and for them the ideal outcome is a file closed early regardless of who was right. A letter costs little, while a lawsuit costs a lot.
The Swiss rule makes the triangle almost geometric. Article 167 of the Insurance Supervision Ordinance guarantees the insured a free choice of lawyer in two cases only, when a representative is needed for judicial or administrative proceedings and when there is a conflict of interest. Even then the insurer can reject the name put forward, at which point the insured proposes three others and one of them must be accepted. Everything that happens before proceedings begin can stay in the hands of the insurer's own legal staff, so the search for an agreement takes place exactly in the phase where the client does not choose who represents him. Germany transposed the same European directive differently. There the insurer may not handle the policyholder's interests directly with its own staff and the right to choose a lawyer applies before any court is involved. Hardly any policyholder knows about this difference when signing, just as hardly anyone reads the general terms until the day they need them.
In fairness to insurers the structure has its own logic. A policyholder who does not pay legal costs out of his own pocket has no brake on fighting over everything and the insurer's control of the early phase also filters out cases that should never have started. The trouble is that this filter has only one economically rational outcome, a quick closure, and that outcome ends up imposing itself even where a genuine negotiation would have produced something better for the policyholder.
It would be wrong to conclude that bargaining should always be avoided. When the relationship between the parties is over and the only thing at stake really is a sum of money, as with an insurer after a minor car accident, hunting for hidden interests is a waste of time and a quick compromise is the right call. The mistake is to use it out of habit when the parties will have to keep working together, or when the table holds more than appears in the lawyers' letters. In the merger I started with it did, and all it took was someone sitting down to look for it.