Reactive Devaluation: Why Their Offer Looks Worse

An offer carries two kinds of information. The first is the terms, which can be read and compared. The second is the fact that someone chose to offer them, which is information about that person's interests. Negotiators use the second kind, and the inference is reasonable: a party that proposes something probably expects to do well under it. Reactive devaluation is what happens when that inference runs past the evidence and changes how the terms are read, so the same four points look generous from one author and thin from another.

The effect has been produced in two ways, and the second is the interesting one. When the source is an adversary, devaluation might simply be the sensible inference above. But the Stanford divestment studies removed the adversary. Students read about two compromise plans for the university's investments and rated which was the bigger concession; the only thing that varied was which plan the university was said to be on the verge of adopting. Whichever plan was described as about to be offered was rated as the smaller concession. Nobody regarded the trustees as an enemy, and the offer carried the same information for both plans, so what moved was the reading of the terms.

Later work looked at that reading directly. Maoz, Ward, Katz and Ross reported in 2002 that differences in how people construed the terms, what each clause would mean in practice, were part of how the attribution did its work. Any real proposal leaves room for interpretation, and the author's identity supplies it.

The consequence for bargaining is specific. Concessions are the currency of negotiation, and this effect devalues the currency on delivery. A side that gives up something real gets less credit than it paid, concludes that goodwill is not reciprocated, and offers less next time. Both sides can then conclude that the other is not serious while a deal that would improve both positions sits on the table. Lee Ross, who named the barrier, was careful about the normative point: discounting an adversary's offer is not in itself irrational. What the studies show is that the discount is applied to the terms, is applied even when the source is not an adversary, and is large enough to block agreements that both sides would prefer to the status quo.

The first evidence came from a sidewalk survey conducted in 1986 by Constance Stillinger, Michael Epelbaum, Dacher Keltner and Lee Ross. The percentages below are the ones Ross printed in his 1995 chapter on reactive devaluation in the volume Barriers to Conflict Resolution; Ross and Stillinger set out the wider argument in Negotiation Journal in 1991. Respondents evaluated a sweeping nuclear disarmament plan: an immediate 50 percent cut in long-range strategic weapons, then further cuts over the next decade and a half until such weapons had gone from both arsenals. The plan was real and had been proposed shortly before by the Soviet leader Mikhail Gorbachev, to little effect.

The survey varied who the plan was said to come from. Attributed to President Reagan, 90 percent of respondents judged it favorable to the United States or evenhanded. Attributed to a group of unnamed strategy analysts, 80 percent did. Attributed to Gorbachev, only 44 percent did. Only responses from people hearing of the plan for the first time were analyzed, and respondents had no strategic reason to misstate their views.

The same group then used a campus controversy over whether Stanford should divest from companies doing business in South Africa, reported in the same chapter. Students read two compromise plans: one divesting immediately from firms tied to the South African military and police, the other creating a monitoring committee with full divestment promised in two years if progress was inadequate. When the university was said to be about to adopt the monitoring plan, 85 percent of students ranked immediate specific divestment as the bigger concession. When the university was said to be about to adopt specific divestment, only 40 percent ranked it as the bigger concession. With neither plan described as imminent, the figure was 69 percent, between the two. A later survey, run before and after the university announced a plan resembling specific divestment, found students' ratings of its significance fell once it was announced.

Maoz, Ward, Katz and Ross took the question to a live conflict in 2002. In the first of three studies, Israeli Jewish respondents rated an actual Israeli-authored peace plan less favorably when it was attributed to the Palestinians than to their own government. In the second, both Israeli Jewish and Israeli Arab respondents devalued a plan ascribed to the other side. The third examined how the two groups construed the plan's terms.

The rows below are illustrations, not findings from the studies cited on this page.

The counters all work by separating the terms from the sender.

• Write down what you would accept before any offer arrives. A stated position is the standard you can hold a later offer against.
• Score the offer blind. Strip the names, or have a colleague who does not know the source rate the terms.
• Ask the question in reverse: if your side had written this, how would you be describing it?
• Count the concession in units. How many days, dollars or seats changed hands compared with the last position on the table? A number is harder to devalue than an impression.
• Name the ambiguity instead of resolving it. Where a clause could mean two things, ask what it means rather than assuming the reading that suits your view of the author.
• If you are the one offering, expect the discount. Put the size of your concession in writing, in the other side's own terms, and ask them to say what it is worth to them.

Reactive devaluation is a change in how terms are valued because of who offered them. Several neighbors describe different responses.

• It is not the not-invented-here syndrome, the preference for solutions developed in house. That is about the origin of an idea inside an organization rather than about a concession in a dispute, and it applies to ideas nobody is bargaining over.
• It is not reactance, the push against a constraint on your freedom. Reactance raises the appeal of the forbidden option; reactive devaluation lowers the appeal of the offered one.
• It is not naive cynicism, the expectation that other people are more self-interested than they are. That is a belief about motives in general; reactive devaluation is a shift in the reading of specific terms.
• It is not the sensible discount a negotiator applies to an adversary's offer. The studies controlled for that: the effect appeared when the source was the university rather than an enemy, and when the only change was whether a plan had been put on the table.