The Power of Alternatives in Negotiation

Introduction

In our last lesson, we explored the Rubinstein bargaining model and established a crucial principle: patience is power. We saw that in a negotiation where the only alternative to agreement is perpetual disagreement, a player's share of the pie is determined by their own patience and that of their opponent.

However, real-world negotiations rarely happen in such a vacuum. Often, one or both parties have a "Plan B"—an alternative they can pursue if the negotiation fails. This is known as an outside option. Today, we will extend our analysis to account for this critical feature of strategic interactions. Our goal is to analyze how a player's outside options influence their bargaining power and the final agreement.

We will discover that the mere existence of an outside option isn't enough; its value relative to the negotiated outcome is what truly matters. This leads to a powerful insight known as the Outside Option Principle.

What is an Outside Option?

An outside option is the value a player can receive by permanently abandoning the current negotiation. Think of it as your best alternative to a negotiated agreement (BATNA). For instance, if you're negotiating a salary with a potential employer, a job offer from another company is your outside option.

To build a clear intuition, let's watch a short video that frames this concept.

Bargaining 101 (#1): Introduction (The Bargaining Problem)

This video by William Spaniel introduces the concept of outside options and the 'bargaining range' using a simple salary negotiation scenario.

Please watch from the beginning to 01:24, and then from 02:07 to 03:04. Pay close attention to how the outside offer of $20/hour sets a minimum acceptable wage for the employee, defining one end of the bargaining range.

As the video illustrates, an outside option establishes a reservation value. A rational player will not accept any offer that is worse than their outside option, as they could simply walk away and take the alternative. This fundamentally changes the strategic landscape.

The Outside Option Principle (OOP)

How do we formally incorporate outside options into the Rubinstein model? We modify the game: after a player rejects an offer, they have two choices:

Make a counteroffer in the next period (as before).
Quit the negotiation and take their outside option.

This leads to a simple yet profound strategic insight: an outside option only affects the outcome if it is credible. A threat to walk away is only credible if the outside option is more valuable than what the player would get by continuing the negotiation. This is the Outside Option Principle.

Let's formalize this. For simplicity, assume two players with the same discount factor, $δ$ . From our last lesson, we know that if Player 1 makes the first offer, the unique Subgame Perfect Equilibrium (SPE) payoffs in the standard Rubinstein game are:

(x_{1}^{*}, x_{2}^{*}) = (\frac{1}{1 + δ}, \frac{δ}{1 + δ})

Now, let's give Player 2 an outside option with a value of $o_{2}$ .

Case 1: The outside option is not credible.
If Player 2's outside option is less than or equal to what she would get from the negotiation anyway ( $o_{2} \leq \frac{δ}{1 + δ}$ ), her threat to walk away is not credible. Why would she take $o_{2}$ when she can get a better deal by staying? In this case, the outside option is irrelevant, and the outcome remains the standard Rubinstein split.
Case 2: The outside option is credible.
If Player 2's outside option is greater than her standard share ( $o_{2} > \frac{δ}{1 + δ}$ ), her threat to walk away is credible. She will not accept any offer less than $o_{2}$ . Player 1, knowing this, must offer Player 2 at least $o_{2}$ to prevent her from leaving. As a profit-maximizer, Player 1 will offer exactly $o_{2}$ . The agreement is reached immediately with payoffs $(1 - o_{2}, o_{2})$ . Here, the outside option determines the outcome.

This principle is sometimes called the "deal-me-out" principle, which is contrasted with a more naive "split-the-difference" approach. The following reading from a seminal paper on this topic explains the distinction.

[PDF] An outside option experiment - UCL Discovery

This paper by Binmore, Shaked, and Sutton was one of the first to experimentally test the Outside Option Principle. The introduction clearly lays out the two competing theories.

Please read the 'Introduction' section (pages 753-757). Focus on the distinction between the 'split-the-difference' prediction (Figure I) and the 'deal-me-out' prediction (Figure II), which corresponds to the Outside Option Principle.

A Formal Proof

Given your interest in mathematical rigor, let's examine the formal proof. The appendix of the paper you just read provides a concise derivation for the unique SPE payoffs in a bargaining game with an outside option. The proof uses the concepts of the supremum (least upper bound) and infimum (greatest lower bound) of the set of possible equilibrium payoffs.

[PDF] An outside option experiment - UCL Discovery

This appendix provides the formal proof of the Outside Option Principle. It's a classic piece of game-theoretic reasoning.

Read 'APPENDIX 1' on page 767. I'll walk you through the logic below.

Let's break down the logic of the proof:

Setup: The proof defines $M_{1}$ and $m_{1}$ as the supremum and infimum of Player 1's equilibrium payoffs when he moves first. $M_{2}$ and $m_{2}$ are the same for Player 2 when she moves first. Player 2 has an outside option of value $s$ .
Key Inequalities:
1. $m_{1} \geq 1 - max {δ M_{2}, s}$ : Player 1's worst possible payoff ( $m_{1}$ ) must be at least what he gets by offering Player 2 just enough to make her accept. Player 2 will accept any offer greater than the best she can get by refusing, which is either her outside option ( $s$ ) or the discounted value of her best possible outcome when she gets to propose next ( $δ M_{2}$ ).
2. The other inequalities, $1 - M_{1} \geq max {δ m_{2}, s}$ and the two for the companion game, follow similar logic.
Case Analysis: The proof then hinges on comparing the outside option $s$ to the value of continuing the game.
- Case (a): $s \leq δ m_{2}$ . The outside option is worse than even the worst possible continuation value. The term $max {δ M_{2}, s}$ simplifies to $δ M_{2}$ , and the game behaves exactly like the standard Rubinstein model, yielding the $\frac{1}{1 + δ}$ split. The outside option is ineffective.
- Case (c): $s > δ M_{2}$ . The outside option is better than even the best possible continuation value. The term $max {δ M_{2}, s}$ simplifies to $s$ . The inequalities then solve to show Player 1's payoff is $1 - s$ . The outside option is effective and determines the outcome.

This elegant proof solidifies the core principle: the outside option must be weighed against the continuation value of the bargaining game itself.

Impact of Outside Options on Bargaining Outcomes — This graph illustrates the Outside Option Principle. The horizontal axis represents a parameter 'T' related to the bargaining horizon (and thus patience), while the vertical axis represents Player 2's outside option, 'b'. The curve $b = (T-1)/(2T)$ represents the threshold where the outside option becomes credible. Below the curve, the option is 'ineffective', and the parties split the surplus. Above the curve, the option is 'effective', and Player 2's share is determined by their outside option.

When Does the Outside Option Principle Fail?

So far, we've assumed a specific structure where a player can opt out after rejecting an offer (i.e., as a responder). But what if the rules were different? Does it matter when a player can exercise their outside option? This question leads to deeper, more nuanced results.

A survey paper by Antoni Cunyat explores exactly this. It turns out that the clean logic of the Outside Option Principle relies heavily on the underlying rules of the game.

[PDF] Strategic Negotiations and Outside Options

This paper explores how the bargaining outcome changes depending on the rules for exercising outside options. It shows that the simple OOP doesn't always hold.

Read the introduction to Section 3 ('Is it so important who has the possibility of opting out?') on page 9, and then read Section 3.1 ('Opting out freely') on pages 10-14. This is a more advanced text, so focus on the main result: when players can opt out at any time (even as proposers), the OOP can fail, and multiple equilibria may exist.

Let's unpack the key finding from the reading:

Opting out as a Responder (The Standard Case): When a player can only opt out after receiving an offer, the Outside Option Principle holds. The proposer can always make a last-minute offer to the responder that is slightly better than their outside option, thereby extracting all the remaining surplus.
Opting out as a Proposer: Things change dramatically if a player can also opt out when it is their turn to make an offer. This ability transforms their proposal into a powerful ultimatum: "accept my offer, or I walk away and we both get our outside option payoffs."

The paper shows that if both players can opt out at any time (as proposer or responder), even a small, supposedly "non-credible" outside option can destabilize the unique Rubinstein equilibrium. The threat to exercise it as a proposer becomes credible because of the bargaining power it confers in that specific subgame. This can lead to a range of possible equilibrium outcomes, destroying the clean prediction of the basic model.

This is a fantastic example of how sensitive game-theoretic predictions can be to the "rules of the game." It also speaks to your interest in using game theory for planning in a chaotic world—understanding the precise rules of engagement is paramount.

Conclusion

In this lesson, we integrated the realistic feature of outside options into our bargaining model. We moved beyond the simple notion that any outside option increases bargaining power and developed a more sophisticated, strategic understanding.

Key Takeaways:

Reservation Value: An outside option sets a floor on what a player will accept in a negotiation.
The Outside Option Principle (OOP): An outside option only influences the bargaining outcome if it is credible, meaning its value is greater than what the player would receive in the equilibrium of the negotiation without the option. If it is not credible, it is strategically irrelevant.
Rules Matter: The validity of the OOP itself depends on the rules of the negotiation. If players can exercise their outside option not just as responders but also as proposers, the strategic dynamics become far more complex, potentially leading to multiple equilibria where even "bad" outside options have an impact.

You have now completed the module on Bargaining and Negotiation. You have mastered the foundational Rubinstein model, analyzed the power of patience, and understood the strategic impact of outside options.

In our next lesson, we will pivot to a different branch of game theory that draws inspiration from biology and population dynamics. We will begin the module on Evolutionary Game Theory and our first goal will be to define an Evolutionarily Stable Strategy (ESS) and its mathematical conditions. This will give us tools to analyze situations where strategies are not chosen by fully rational players but rather evolve over time in a large population.

Recommended resources