An auction makes the institution behind a price unusually visible. A seller announces what can be won, who may bid, how the winner will be chosen and what the winner must pay. Change one of those rules and the offers can change, even when the object and the people who want it remain the same. The surplus is therefore being divided by more than the intensity of the bidders' desire. It is being divided by a procedure that gives that desire a reason to reveal or conceal itself. The question for this plate is a precise one: can a selling rule place a good with the participant who values it most, while leaving each participant to act in his own interest? William Vickrey's second-price auction supplies a striking answer under stated conditions. Its distinction between the bid that wins and the price that is paid is the small hinge on which the result turns.
Consider one indivisible object, a seller whose value of retaining it is zero, and three buyers whose values are a hundred, eighty and fifty. Each knows his own value, can pay it, and cares only about obtaining the object and the money he keeps. There are no effects on outsiders, no side agreements and no gain from another buyer's payment. In a sealed second-price auction, each submits an offer without seeing the others; the highest offer wins, but the winner pays the second-highest. If the buyers report their values, the first receives the object for eighty and retains a surplus of twenty. The seller receives eighty. Together their monetary gains are a hundred, the value of the object to its new holder less its value to the seller. The runner-up's eighty divides that total; it does not create it. A price of ninety would divide the same total differently if the same buyer still received the object. The allocation and the payment are two parts of the mechanism, with different work to do.1
Truthful bidding is a weakly dominant strategy in this setting because a bidder's offer decides whether he wins, while the strongest rival's offer fixes what winning costs. Hold the rival offers still and call the highest of them eighty. The buyer whose value is a hundred wants to win at that price. Reporting a hundred secures the gain of twenty; reporting seventy loses it; reporting a hundred and thirty gives no larger gain. Now hold the highest rival offer at a hundred and ten instead. The same buyer wants to lose, since winning would cost ten more than the object is worth to him. A truthful offer loses; an offer of a hundred and thirty can make him win at a loss. Reporting his value thus gets the favorable purchases and avoids the unfavorable ones. Some other offers produce exactly the same outcome, which is why dominance is weak rather than strict. The proof does not require guessing a probability distribution for the rivals' offers. It requires that the buyer's own value and the announced payment rule remain what the comparison assumes.2
A first-price auction changes the incentive because the winner pays his own offer. Bidding the full hundred would leave a buyer with value a hundred no gain whenever he won; a lower bid leaves a gain but risks losing the object. The choice now depends on the offers he expects from others. This does not mean that a first-price rule must always put the object in the wrong hands, or that the second-price rule must collect less revenue. Take two risk-neutral buyers with values drawn independently and uniformly between zero and a hundred, each knowing that distribution. In the symmetric first-price equilibrium, each bids half his value. The highest-value buyer still wins because halving preserves their order. The expected highest value is two hundred divided by three, so the expected winning bid is a hundred divided by three. In the truthful second-price auction, the price is the lower of the two values, whose expectation is also a hundred divided by three. Equal average revenue here is a result of the whole specified model. It neither makes the realized prices identical nor promises equivalence for every set of bidders.3
The seller's objective can also change the division and the allocation. A reserve of ninety in the first illustration would make the truthful highest-value buyer pay ninety rather than eighty, leaving him ten. But a reserve of a hundred and ten would prevent the sale entirely, despite the hundred of value available from transferring the object to him. A seller who does not know the buyers' values must weigh a possible higher payment against the possibility of no sale. Maximizing expected revenue is therefore a different objective from ensuring that every beneficial transfer occurs. The difference also clarifies what the winner's retained surplus means. The second-price rule leaves a gap to the next offer under its conditions; that gap is part of how the mechanism works, rather than proof that the seller overlooked an obvious additional charge. If a seller already knew every value, the informational problem would be different. The earlier plate, The Sorting, examines the extraction such knowledge can make possible.4
A good rule also needs a setting in which its promises are credible. The buyers must believe that the competing bids are genuine and that the seller will apply the announced payment rule. Vickrey discusses this administrative problem alongside the mathematics, including the danger of collusion and fabricated competing offers. Information can present a different difficulty. If the object will be resold in a common market, the evidence held by other bidders may help determine what it is worth; a buyer's initial estimate is then not the fixed private value used in the proof. The separate Auction sets this beside broader choices in market design. Budgets, complementarities between several objects and effects on people outside the sale can likewise change the allocation question. Even in the simple model, highest monetary willingness to pay is not a direct ranking of human need. The mechanism elicits a kind of value under its rules and constraints. Deciding whether that is the value the institution ought to maximize remains a further task.5
The bidding shows how a price can carry information because the rules make providing that information worthwhile. In the simple second-price auction, the offer can reveal a private value without making the bidder pay that entire value. The winning buyer keeps the difference to the next offer, the seller receives the payment, and the object reaches its highest-valued use among the participants under the model's conditions. These are achievements of a specified institution, not properties of every event called an auction. They also leave us with a distinction that matters beyond the auction room. Money paid for the prize changes hands; the time and resources spent obtaining it may be consumed. A procedure can allocate a good well while demanding costly effort, and a contest can invite effort that changes only who takes an existing gain. The next plate, The Scramble, follows that possibility: the surplus as a prize whose pursuit can eat into the advantage the winner hoped to keep.