“Split the Pie” – a Tasty Take on Negotiation

Negotiation is the most common form of dispute resolution.

I recently heard that about 50,000 actions are started in the Ontario Superior Court of Justice each year. About 2,000 cases go to trial. The rest are resolved some other way – mainly negotiated settlements.

I don’t know how accurate those numbers are, but that’s an awful lot of negotiation.

The difficulty, in many cases, is that parties (and counsel) don’t really understand what they’re negotiating about. It’s not the merits of the dispute. It’s not even the dollars claimed on each side.

It’s the difference in value between a deal and no deal.

That brings me to a very interesting new book on negotiation: Split the Pie by Barry Nalebuff. Subtitled “A Radical New Way to Negotiate”, the book is based on the author’s negotiation course at Yale University where he has taught for more than 30 years.

He starts with a simple thought experiment:

Students Alice and Bob are at a local pizza joint in New Haven. They’re both hungry but only have enough money for a smallish 6-slice pie. Alice has twice as much cash than Bob, so they agree to split it 4 slices for Alice and 2 for Bob. Pepe, the owner, offers them a deal – a big 12-slice pizza for the same price, but only if they can agree how to divide it up. Otherwise, they get the smaller one.

How should they split the pizza?

Power/Resources: Alice has twice as much cash as Bob and gets twice as many slices if there’s no deal. She wants 8 pieces; Bob can have 4.

Fairness: They’re both equally hungry and they’ve been offered a free bonus. They should each get half the pie – 6 each. (Alice should be happy with two extra pieces and probably can’t eat more anyway…)

These are common positional and interest-based approaches to negotiation. Nalebuff argues that they’re both wrong. Alice should get 7 pieces; Bob should get 5.

This flows from the realization that they’re not negotiating over how to split 12 slices. The negotiating “pie” is actually 6 slices – the extra slices they get if they make a deal. They should agree to “split the pie” equally – 3 extra slices – because they are equally necessary to make a deal and get the bonus.

It’s a very simple hypothetical, but Nalebuff works through the logic of each argument and concludes that splitting the pie equally is the only principled approach that works in every case.

Take for example, the case where Bob only has enough cash for one slice and Alice has enough for 5. Should the split be 10-2 for the bigger pie? Should they both get 6 pieces regardless of how much they started with? Or should they still split the extra pie 50-50? No, they should still get 3 extra pieces (leading to an 8/4 spit).

The critical point is both sides are equally necessary to make a deal and grow the pie, so both sides should get an equal benefit.

The examples quickly get much more complex. They include:

• Siblings pooling their money to get a better interest rate on an investment.

Alice has $10,000 to invest in a GIC; Bob has $20,000. Their bank is offering 3% interest up to $20,000; 4% on $20,000 or more. Alice can make an extra $100 if the pool their money. Bob won’t get anything extra. How should they split the extra interest?
“Splitting the Pie” says the pie (joint benefit) is the extra 1% on Alice’s share ($100) so they should get $50 each. They are each $50 better off than if they don’t work together.

• Splitting the cost of municipal services to a subdivision.

Developers Alpha and Beta are building new subdivisions. They both need new municipal water and sewer services. The lines for Beta will run straight past Alpha’s property. The city is charging $xx a meter to put in the services; they each have 1000 meters of frontage. How much should Apha and Beta pay?

Beta argues that Alpha needs the services anyway, so Beta should only pay for the extra distance to their own property. Alpha says that’s not fair; Beta would have had to pay for the full 2000 meters if they go it alone. They should both save money if they share the same pipe.

The “Split the Pie” principle says the cost for the first section should be split evenly and Beta should pay the full cost for the last section. This principle can be extended to any number of parties in the deal. The cost saving for shared sections is shared equally among all those who benefit.

Nalebuff acknowledges that the calculations (figuring out the size of the pie) can become very complex, but the same principle can apply to a wide range of situations.

How should you split the airfare for a multi-leg business trip among several clients when the combined cost of all the legs is less than multiple round trips? Nalebuff argues that the saving should be split equally, not proportionately based on the cost of each individual flight.

Or how should electric vehicle manufacturers split the cost of installing charging infrastructure throughout Europe or North America when each will benefit differently? (This is an actual case study from the European Union where the larger companies grabbed most of the shared cost benefits for themselves. The book makes the argument for why relative size or power shouldn’t matter.)

Nalebuff also deals with situations where the size of the pie is uncertain. The parties can agree to split the pie once it is baked. They just need to agree on how to value it.

Split the Pie also offers a principled approach to dispute settlement negotiations. The challenge is that, when resolving a dispute, one or more of the parties is negotiating over an inevitable loss rather than a potential gain. The psychology is quite different. It’s much harder to accept a loss – even a partial loss – than it is to share a gain. For example:

• Nalebuff’s own negotiations with a cybersquatter over a domain name.

The squatter wanted $2500 for the domain name. Nalebuff found it would cost only $1300 to use the ICANN dispute process to get the domain back. They haggled by email for a while. Then he pointed out the negotiation wasn’t really about the arbitrary demand for $2500. It was about the $1300 ICANN fee. That’s what it would cost to get the domain back if there was no deal. He proposed to “split the pie” – the guy who registered the name gets $650; Nalebuff saves $650 he would otherwise have to pay ICANN. Despite some more push-back, they agreed on $650 because there was no effective counterargument.

(In this example, Nalebuff fails to take into account the possibility, however remote, that he might lose the domain name arbitration. Victory may be highly likely, but it’s not guaranteed. In litigation and arbitration, nothing is absolutely certain.)

• Damages for cancelling a contract.

The parties can agree to minimize their joint loss by agreeing to “split the pie” (which is anything better than the worst-case scenario).

One example is someone breaking a lease to take a job in another city. The tenant’s potential loss is the rent for the rest of the lease (just leave the place empty if they can’t sublet it or make a deal with the landlord). The landlord benefits from re-renting to a long-term tenant rather than allowing a sublet. The landlord also has an incentive to try to rent it again quickly if they can collect double rent for a few months. But no incentive if all the money goes to the tenant.

Nalebuff’s answer is to agree to split the pie. The size of the pie is the potential double rent the unit could generate (especially if the rental market is tight and it can be rented quickly). Landlord keeps half and rebates half to the departing tenant.

The same idea applies in other situations, such as terminating an employment agreement. Splitting the pie is essentially what happens when an employer offers salary continuance for the legal notice period, with a lump sum equal to half the remaining notice period if the employee gets another job.

In all of these negotiations, each side may be tempted to try to negotiate a more favourable split but Split the Pie offers a principled argument for why it should be 50/50, regardless of negotiating power, who’s “at fault” or which side cares more about the outcome. Both sides have to cooperate to maximize the size of the pie, so they should share the benefit equally.

The main flaw in this approach to negotiation, I think, is that Nalebuff assumes that the parties are all rational economic actors. (After all, he is a business school professor.)

The reality of negotiating a dispute settlement is that people don’t always act rationally. They may want retribution or punishment. They may simply reject any deal where the other side benefits.

“Splitting the pie” means both sides benefit equally from any gain – or reduced loss. That may be a tough sell in settlement negotiations. But it is a principled approach. And one that lets the parties focus on negotiating ways to increase the size of the pie rather than arguing over who should get the bigger slice.

Still skeptical? Read the book.

Split the Pie by Barry Nalebuff
2022, Harper Business, New York