Articles & Posts

Why are two sides who read exactly the same file both sure they will win?

Three answers from the research

Most people think a dispute doesn't settle because someone is hiding information, or because the other side is simply greedy.
The research shows something more troubling:
even when both sides know exactly the same things, each of them sees a different case.

1. What happens when two people read the same 27 pages?

In 1993, four researchers (Loewenstein, Issacharoff, Camerer and Babcock) split 160 students into pairs: one "plaintiff", one "defendant".
The case: a real accident from Texas, a motorcyclist who sued a driver for $100,000.
Both received the same 27 pages: testimony, a police report, maps.
Each one secretly guessed what the real judge had awarded, with a prize for getting it right.
The "plaintiffs" guessed, on average, $14,527 more than the "defendants".
Why?
Nobody lied. The guess was secret, and there was a prize for accuracy.
Each of them simply read the same pages from their own "chair".
And in the pairs that failed to settle, the gap was three times larger.

2. When exactly does the mind start to tilt?

In 1995 the researchers changed one detail: the order of things.
Some participants knew their role before reading the case. Among them, 28% of the pairs failed to agree and left the decision to the judge.
Others learned their role only after reading and estimating. Among them, only 6% failed to settle.
What does this tell us?
The bias is not created during the argument. It is created already while reading.
And the most amusing part: when the participants were told about this bias, they concluded that the other side was probably biased. Their own estimate they did not change.

3. What does it cost in the real world?

In 2008, Kiser, Asher and McShane examined 2,054 civil cases in California in which the parties rejected a settlement offer and went all the way.
In 61% of the cases, the plaintiffs ended up with an amount no higher than the offer they had rejected.
The defendants were wrong less often, in 24% of the cases, but their mistake cost on average $1.14 million.
And it is not only the clients: in a 2010 study, 481 American lawyers estimated in advance their chances of reaching a goal they had set in a case. On average they were overconfident, and years of experience did not improve their accuracy.

4. Bonus

There is one exercise that worked in the lab.
Babcock and her colleagues asked participants, before negotiating, to write a list of the weaknesses in their own case.
The gap between the sides dropped from $21,783 to $4,674, and the share of pairs that failed to agree fell from 35% to 4%.
So the next time you are 100% sure you are right, against a partner, a supplier or a tenant, take a sheet of paper and write down three reasons why you might not win.
(And no, "the other side just doesn't get it" doesn't count.)

Bottom line: both sides read the same file.
But probably only one of them will like the judgment.

Meanwhile, I'm off to try to convince the CEO of a company, my client, that merely talking about a settlement or mediation with the other side does not immediately signal weakness. Have a good week!

All articles and posts