The Price of Candor

Why the information that creates the deal can also change its price

By Glenn C Nunes

I. October 1

A buyer needs a widget delivered by October 1.

The seller's standard schedule delivers in November. For $20 of overtime and rearranged production, the seller could deliver in September. On-time delivery is worth an additional $200 to the buyer.

Run the negotiation twice.

In the first version, the buyer says nothing about timing. Why would he? Timing is his problem, and revealing a need feels like revealing a weakness. The parties negotiate price, agree on a number, and part satisfied. The widget arrives in November. The buyer quietly absorbs the $200 loss.

Nobody behaved foolishly. The seller never knew speed was worth anything. The buyer never knew speed was available for $20. A trade that would have cost twenty dollars and produced two hundred died in silence.

That is the case for candor. It is a good case. An expedite worth $180 in new value went undiscovered because the one fact that would have revealed it stayed private.

Now run it again.

This time the buyer explains everything: October 1 is essential, no other supplier can perform in time, the marketing campaign has already launched, and a late widget costs him $200.

The seller now knows two things.

She knows how to solve the buyer's problem: expedite for $20.

She also knows what the solution may be worth: as much as $200.

So what should the expedite cost? Twenty-five dollars? One hundred? A hundred and ninety-five? Every price in that range beats the alternative for both sides — and the buyer has just disclosed the top of it. The seller already knew the bottom; it was her own cost. What she lacked was any idea of the ceiling, and the buyer has handed it to her.

Notice what happened. One fact found the trade: October 1 matters. The other facts priced the buyer's need for it: no alternatives, campaign launched, $200 at stake. The buyer did not make one disclosure. He made several, and they did different work.

In the first version, silence destroyed $180 of value. In the second, candor put most of it up for grabs.

That is the problem this essay is about.

One clarification before going further, because the word in the title can mislead. Candor here means the voluntary revelation of private bargaining information — what your side values, what it can pay, what pressures it faces. It does not mean truthfulness, which is not optional. Nothing in this essay excuses a false statement, and nothing in it excuses withholding information that law, contract, or professional duty requires a party to disclose. Those obligations come first. The strategic question — what to volunteer, when, and in what form — begins only after they are satisfied.

II. The negotiator's dilemma

The tension in the October 1 story has a name.

In 1986, David Lax and James Sebenius called it the Negotiator's Dilemma.1 Creating value rewards openness: the trades that make a deal worth more are discovered by comparing interests, priorities, capabilities, and costs, and comparison requires learning — someone must come to know something the other side knows. Claiming value rewards guardedness: the party who reveals what an issue is worth to him has made it easier for the other side to estimate what it can charge. Every negotiator needs both, and the behaviors pull against each other. Openness exposes you. Guardedness blinds you both.

The first essay in this series argued that negotiation culture elevated value creation while treating value claiming as crude. The second argued that creating more value does not determine how that value will be divided — the bigger pie still has to be cut.

This essay is about the mechanism that connects them.

The reason creating and claiming cannot be cleanly separated — the reason "first we'll expand the pie together, then we'll divide it" never quite works — is that the two tasks run on the same fuel. The information that expands the pie is information about what the parties value. And the more precisely a party reveals what it values, the easier it becomes for the other side to estimate what it will pay.

The buyer cannot tell the seller where the value is without telling her something about whose it is.

III. The two jobs a fact can do

If neither rule survives — not "be transparent," not "reveal nothing" — then what's needed is a way to tell disclosures apart.

Information does many things in a negotiation. It can build trust or spend it, establish credibility, coordinate performance, satisfy a duty, signal competence or desperation. This essay examines the two economic functions at the center of the dilemma, and before revealing a fact it is worth asking at least these two questions:2

Does it reveal or enable a deal the parties did not know was available?

Some information surfaces a trade, permits a different performance, uncovers a cheaper way to bear a risk, or prevents an impasse neither side wants. "We need delivery by October 1" is this kind of fact. The seller could always expedite — that capability was hers all along. What she could not do was know that anyone would pay for it. The disclosure did not change her powers. It revealed a deal worth making.

Or does it mainly improve the other side's estimate of what it can charge?

Some information surfaces no new deal. It reveals a maximum, a weak alternative, an internal pressure, a cost of delay. "A late widget costs us $200" is this kind of fact. The seller learns nothing that lets her do anything new. She learns what the thing she was already going to do is worth to the person buying it.

Call the first kind productive information and the second kind pricing information. Productive information tells the other side where the value is. Pricing information tells them how much of it you may surrender.

The distinction is diagnostic, not clean. Plenty of facts do both jobs at once — even "October 1 matters," the most productive disclosure in the story, whispers urgency. That is exactly what makes the dilemma a dilemma rather than a sorting exercise. But most negotiators never ask the question at all. They treat disclosure as a matter of temperament: open people share, careful people don't. The temperaments are both wrong. The useful question is not how much to reveal. It is which job each fact will do once it is loose.

IV. The fact and the reason

Watch the test work on the October 1 problem.

"We need delivery by October 1" — mostly productive. It puts the expedite on the table. Without it, there is no trade to discuss.

"Because our campaign has launched, no other supplier can perform, and delay costs us $200" — mostly pricing. The seller's capabilities haven't changed. Her estimate of the buyer's dependence has.

The buyer did not face one disclosure decision. He faced two, with different answers. The constraint may well have been worth revealing. The explanation of his dependence was considerably more expensive.

This division shows up wherever constraints appear. A deadline is the cleanest case. Consider a negotiation that genuinely ends on Friday — the financing window closes, for everyone. Concealed, that deadline pressures only the party who knows about it: his clock runs while the other side proceeds at its leisure. Spoken aloud — "this must close by Friday" — it becomes a shared fact that may hurry both parties, because delay now visibly threatens a deal they both want. It may do other things too: it can be doubted, or read as desperation, or matter little to a party with strong alternatives.

Now watch what the explanation does. "Because our financing expires Friday" may make the deadline believable — without some reason, "Friday" can sound invented, a pressure tactic to be waited out. But "and we have no replacement" also tells the other side exactly how weak your alternative is. The same explanation strengthens the coordinating message and raises the price of sending it. That is the dilemma in a single sentence: the words that make the constraint credible are the words that expose the dependence behind it.

The same split appears in softer form between an interest and a priority. "Timing matters to us" opens a trade. "Timing matters more than every other term in this contract" tells the other side where concessions may be easiest to obtain. The first is a door; the second is a map.

None of this yields a rule as tidy as "reveal interests, conceal limits." Sometimes the reason is harmless. Sometimes the priority must be shared or the trade cannot be built at all. The point is narrower and more durable: a disclosure is not a unit. Many important disclosures can be divided into the part that surfaces the deal and the part that prices your need for it, and nothing obliges you to deliver both parts at once.

V. Disclosure is not enough

At this point a careful reader may draw the wrong conclusion: guard everything that smells like pricing information, and reveal the productive facts freely, and the problem is managed. The economists supply two corrections, and the first is that revealing is harder than it sounds.

In 1970, George Akerlof published a short paper about used cars.3 Sellers know which cars are sound and which are lemons; buyers cannot tell the difference. So buyers will pay only a price reflecting average quality. But at that price, the owners of the best cars refuse to sell, and here is the detail that matters: it is not that they won't disclose their quality — they would love to. It is that they can't, credibly, because the owner of a lemon can say the same words. Every seller claims a sound car; the claims cancel; the good cars leave the market; average quality falls; and the unraveling feeds itself. Trades that would have benefited both sides never happen — not because anyone chose silence, but because saying it could not make it believed.

Information can perform either of its jobs only to the extent that it is believed. Akerlof's honest seller is willing to reveal everything and still loses the trade, because the buyer cannot distinguish his truth from a lemon owner's identical claim. A statement the other side heavily discounts does little work — productive or pricing — no matter how accurate it is.

Akerlof's own remedies point somewhere useful: not fuller explanation, but warranties, brand names, certification — arrangements that make the claim costly to fake. The answer to a credibility problem is rarely more words. It is a commitment, a verification, or a proposed exchange that a liar would not offer. Hold that thought; it is where the practical techniques in the next section get their power.

The second correction runs deeper. In 1983, Roger Myerson and Mark Satterthwaite examined the simplest possible negotiation — one buyer, one seller, one item, each privately knowing its own valuation.4 Under the paper's assumptions — private and independent valuations, voluntary participation, incentives for truthful reporting, and no outside subsidy — no bargaining procedure can guarantee that the parties trade every time a mutually beneficial trade exists. Some deals that should close, won't. In ordinary bargaining the intuition is familiar, though it is an intuition and not the theorem itself: each side shades — the buyer understates, the seller overstates — and some trades disappear into the gap between the shadings. The theorem's contribution is showing that under those conditions this is not a failure of technique that better negotiators simply avoid. It is built into bargaining with two-sided private information.

Together the two results reframe the problem. Private information creates friction that exhortations to openness cannot eliminate. "Just be transparent" fails twice: the transparency may not be believed, and even honest dealing cannot guarantee every good trade. Candor has a price — the buyer who explains everything hands the seller his ceiling. But secrecy is no safe harbor, and neither is sincerity. The trades that die undiscovered, the truths that can't be distinguished from lies, and the deals that collapse in the gap between shaded positions are all paid from the same account.

Silence is not neutral. It is a bid, and sometimes it loses.

VI. Disclosure by design

Which brings the practical question: how does the October 1 buyer get the expedite without publishing the $200?

The choice was never binary. Between telling everything and telling nothing lies a set of techniques that reveal possible trades while limiting how directly you expose your limits. None of them eliminates the leak — every move in a negotiation communicates something — but they change its size and its form. And the best of them do double duty: they answer the credibility problem at the same time, because they carry information in forms that are costly to fake.

Ask before telling. The buyer's first move need not be a statement. "What do your delivery schedules look like? What would an earlier date require on your end? What would that cost?" Questions can surface the possibility of expedited delivery — and the seller's proposed price for it — before the buyer has explained why he wants it. Asking about speed does signal that speed may matter; a question is never free. But there is a difference between disclosing a topic and disclosing a dependence, and the question discloses only the first.

Make the disclosure conditional. "Speed is extremely important to us" is an announcement, delivered free. "If you can guarantee October 1, we could adjust another term" is an exchange. The seller learns that timing matters — she must, or no trade — but the interest arrives already attached to a proposed payment, as a trade to accept or counter rather than an anxiety to exploit. There is a credibility bonus, too: an interest embedded in an offer is more believable than an interest merely asserted, because the buyer is visibly willing to pay for it.

Let packages do the talking. Offer several bundles that are roughly equivalent from your own point of view: a lower price with November delivery; a middle price with October delivery; a premium with September delivery and narrower warranty protection. Because the packages cost you about the same, the seller's response is informative — it may reveal her relative costs and priorities, since she has no reason to pick one bundle except that it suits her side better. Each party learns the shape of the other's preferences from the pattern of trades, indirectly and mutually, without either stating a walk-away point out loud. The priorities are not secret; they are legible in the choices. But there is a difference between letting the other side infer your ranking from trades you propose and announcing your dependence in a sentence they can quote back to you.

Stage what remains. Disclosure is irreversible. You can always reveal the next fact; you cannot make the other side unlearn the last one. That asymmetry is the argument for sequence. Reveal what the deal requires now; hold what it doesn't; and remember that a fact released to create value in the morning will still be on the table, doing its other job, when price comes up in the afternoon.

None of this is deception, and none of it solves the dilemma. The buyer using every technique has told no lies and may end the deal having disclosed nearly everything. What he has controlled is order and form — which is to say, he has treated his facts as assets and spent them like someone who knows they only spend once.

VII. The orange

Negotiation's most famous parable involves two sisters and one orange: each wanted it, they split it down the middle, and only later discovered that one wanted the fruit and the other wanted the peel. Naming their interests would have divided the orange for them — which is why the story became the standard argument for candor.

But the orange was generous to its owners. Within the parable, each sister's disclosure was purely productive and priced nothing, because neither had any use for the part she gave away.

Real negotiations are rarely so polite. The fact that surfaces the trade often travels with facts that price your need for it, and information does not ask your permission before doing its second job.

Candor has a price. Secrecy has one too. The skill is not openness, and it is not concealment.

It is knowing what the next fact will do.

Notes

1. David A. Lax & James K. Sebenius, The Manager as Negotiator: Bargaining for Cooperation and Competitive Gain (1986). Lax and Sebenius present the tension between value-creating and value-claiming behavior as the central strategic problem of negotiation.

2. The productive/pricing distinction, and the suggestion of applying it fact by fact rather than to categories of information, are my own analytical devices rather than claims made or tested by the cited authors. They are the two economic functions examined in this essay, not the only functions information performs.

3. George A. Akerlof, The Market for "Lemons": Quality Uncertainty and the Market Mechanism, 84 Q.J. Econ. 488 (1970). Akerlof also identifies counteracting institutions — guarantees, brand names, chains, licensing, and certification — that make quality claims credible.

4. Roger B. Myerson & Mark A. Satterthwaite, Efficient Mechanisms for Bilateral Trading, 29 J. Econ. Theory 265 (1983). Stated informally: where one buyer and one seller each privately know their own valuations, no bargaining mechanism satisfying the conditions specified in the paper — including independent private valuations, voluntary participation, incentive-compatible truthful reporting, no external subsidy, and overlapping ranges of possible valuations — can guarantee trade in every case where trade would benefit both parties.