The Art of Negotiation in a Bustling South Asian Market — Epoche B2
Stop Telling Tourists to Offer Half "Whatever the seller says, offer half, then meet somewhere in the middle" is the single most repeated piece of bargaining advice given to visitors to India, Sri Lanka and Nepal, and it is wrong in two separate ways that this essay will take in turn: it is wrong arithmetically, and it is wrong about what a bargain is for. It appears in print, on travel forums, and in the mouths of people who have been in the country three days advising people who have been there for one. It is bad advice, and it is bad in a way that damages the person following it rather more than the person it is aimed at. What the rule assumes, and when it is right The rule rests on two assumptions: that the asking price is a fiction, and that the exchange is a contest with a winner. The first is sometimes true, and it has a name. Charging different prices to different groups of customers for the same good, according to what each group will pay, is price discrimination — a standard and entirely rational strategy for any seller who can tell the groups apart and stop them reselling to each other. A tourist stepping off a coach is trivially identifiable and will not resell, so a stall selling identical printed scarves outside a monument may well quote a foreigner several times the local price. Against that, halving is a rough correction and it works by accident. But the rule is applied indiscriminately, and it is applied hardest by exactly the visitors least equipped to tell one kind of stall from another. The rest of this essay is about what it does everywhere else. The arithmetic, which nobody does Start with the part that requires no fieldwork. The rule's output is a function of the seller's opening number and nothing else, which means the seller controls the result. Amos Tversky and Daniel Kahneman described the underlying mechanism in 1974 as anchoring: when people estimate an unknown quantity after being shown a starting value, their estimates stay close to that value, and they do so even when the starting value is arbitrary and known to be arbitrary. "Offer half" does not merely fall prey to anchoring; it is an instruction to compute the answer from the anchor. Take a good whose fair local price is 1,000 rupees and run the rule against two sellers. Seller's opening quote Buyer offers half Meet in the middle Result versus fair price 3,000 (tourist quote) 1,500 2,250 2.25 times too much 1,000 (honest quote) 500 750 below the seller's cost The first row is the case the rule was invented for, and it fails: halving three thousand leaves you above the real price before the meeting-in-the-middle step pushes you further above it. The second row is worse than failure, because 750 is a price at which an honest seller cannot trade at all. He has three options: refuse, and lose the sale; accept, and lose money; or find something he can sell for 750. What he does is the third, and the next section is about what that looks like. Cardamom, and what you cannot see in the bag The clearest place to watch the third option is a market where quality varies invisibly and price tracks it closely. Consider the spice warehouses of Mattancherry, in Kochi, where the cardamom comes down from the Idukki hills in Kerala. Cardamom is not one commodity. Indian small cardamom is graded chiefly by capsule size and colour, with the size determined mechanically by which sieve mesh the capsules are retained on — the trade grades that carry words like "bold" and "extra bold" are literally statements about millimetres. Grade designations for agricultural produce in India have statutory backing under the Agricultural Produce (Grading and Marking) Act of 1937, and cardamom in particular is regulated by the Spices Board, the statutory body created by the Spices Board Act of 1986, which supervises the licensed auctions through which the crop passes on its way from grower to shop. Those auctions publish their prices by grade. The important consequence for a visitor is this: the "real price" is not a secret held by the trade. It is a published number that differs by grade, and the gap between the top grade and an ordinary one runs to several hundred rupees a kilogram. Now put the halving rule into that shop. If the seller quoted a tourist price, the visitor gets a fair deal by accident. If the seller quoted the real price of the good grade, something else happens, and it is worth setting out formally because it is the heart of the matter. Write $p_H$ for the price at which the seller will part with the high grade and $p_L$ for the price at which he will part with the low one, with $p_H$ the larger. An offer that is at or above $p_L$ but beneath $p_H$ cannot be met with the high grade — not because the seller is dishonest, but because he would be selling below what the auction charged him. It can be met with the low grade. So the offer does not buy the good cardamom cheaply. It selects the poor cardamom, into a bag the visitor will not open until they are home. This is George Akerlof's market for lemons run in miniature. Akerlof's 1970 argument concerned used cars: when the seller knows an item's quality and the buyer does not, the buyer will only pay what an average item is worth; owners of better-than-average items then decline to sell at that price and withdraw; the average quality of what remains falls; the buyer, observing this, offers less again. Economists call the mechanism adverse selection — the price itself selects which quality shows up. The visitor who insists on the price of the poor grade will be sold the poor grade, and the good grade quietly leaves the counter. The rule did not protect anybody. It merely moved the point at which they were overcharged. The second cost: what your opening tells the seller The arithmetic and the adverse selection are both about a single transaction. The subtler cost appears once you notice that a bargain is not only about price. Clifford Geertz spent years watching the bazaar in