#838 1897 · William P. Brown (New York stamp auctioneer) · Auctions & collectibles

A Manhattan stamp dealer told mail bidders their own bid would never set the price they paid

the problem

Absentee bidders sending sealed offers by mail had to guess a price without seeing rivals' bids

background

Stamp collecting boomed in the 1870s-1890s and auctions had multiplied nationwide, but the serious buyers were scattered across the country while the sale rooms sat in New York. Dealers had already started waiving the old 10 percent fee for representing absentee bidders (Brown's own 1878 catalog was the first), yet the payment rule underneath was still first-price: whatever you wrote on your bid slip was what you paid if you won. That left every out-of-town bidder guessing — shade the number too low and lose the lot to whoever guessed higher; write your honest top price and, if no one else came close, pay far more than the lot would have fetched.

The theoretically clean fix already existed and everyone knew it: a live ascending auction, where bidding simply stops the instant your last rival quits, so the winner only ever pays one small step above the runner-up, never their own ceiling. But a live auction requires the parties to actually go back and forth in real time, and a sealed bid arriving once in an envelope cannot do that. Whatever rule a dealer wrote into the catalog had to fake the floor auction's stopping reflex on paper, with no live exchange to fake it in.

what everyone would do

Keep the standard mail 'tender' sale, where the winner pays exactly what they wrote down. It was already the norm — several early mail sales in the 1870s used it — but it forced every bidder to shade below their honest value to avoid overpaying, which cost the dealer revenue from whoever guessed wrong in either direction.

what they saw

In a live floor auction, the winner never actually pays their own ceiling — they pay one increment above whoever dropped out second-to-last, because that's the moment bidding stops. Brown saw that this stopping rule, not the physical presence of rival bidders, was the part doing the work — and that a rule, unlike a room full of people, can be written into a catalog and executed by mail.

the move

Brown's 1897 catalog, Auction Sale No. 2, told bidders their offers were confidential and that the highest bidder would win — but at a price set only one small increment above the second-highest offer received, never at the amount the winner had actually written down. He spelled out a worked numeric example directly in the terms of sale so no one had to take the promise on faith: with a $1.50 reserve and competing bids of $1.60, $2 and $10, the $10 bidder would win the lot for $2.10, 'the lowest price at which he could obtain it if all the competitors were present.'

why it works

Once the price a bidder pays is set by someone else's offer rather than their own, the size of your own bid can no longer cost you money except by determining whether you win at all — so the dominant strategy collapses to simply writing your true maximum. Honest bidding then pulls the dealer's revenue up toward each lot's real value instead of down toward whatever the most cautious credible bidder guessed the market would bear, and it does this for bidders who could never travel to a live sale room in the first place.

the payoff

Bidders could now write their true maximum with nothing to lose — bidding high never cost more than the next-best offer plus a few cents, so Brown captured close to the lot's real value without a single bidder needing to physically attend or haggle. The mechanism spread through the mail-order stamp trade over the next three decades (B.L. Voorhees from 1906, Western Stamp Co. and Toledo Stamp Co. from 1907, Henry Wendt from 1920), and by the 1930s it was the majority rule for stamp mail sales, stated so routinely that dealers stopped bothering to explain it.

where it breaks

The guarantee only holds if bidders trust the stated second-highest bid is real. Lucking-Reiley's own interviews caught a stamp auctioneer admitting he sometimes charged winners their full bid instead of the announced second-price rule when he needed the cash, and collectors described deliberately shading their mail bids to unfamiliar dealers for exactly that reason. Without a reputation mechanism, an audit trail, or (as eBay later added) a public list of the losing bids, the seller can quietly revert to first-price behavior, bidders rationally start shading again, and the whole advantage collapses back into the problem it was built to solve.

what came after

Economist William Vickrey derived the identical rule mathematically in 1961, apparently unaware any stamp dealer had ever run it, and for decades economists credited him as the inventor of the 'second-price auction' — a 1990 paper even asked 'Why Are Vickrey Auctions Rare?' without knowing several hundred small stamp firms were already running one every month. David Lucking-Reiley's 2000 paper stumbled onto the precedent while researching early eBay auctions and restored the history; the same logic now runs eBay's proxy bidding and, in generalized multi-slot form, the ad auctions that price search and social-media advertising at Google and Meta.

references

[1]Vickrey Auctions in Practice: From Nineteenth-Century Philately to Twenty-First-Century E-CommerceJournal of Economic Perspectives (American Economic Association), 2000doi.org
[2]A Survey of Auction Theory (citing Lucking-Reiley's stamp-auction history)Princeton University Press / Journal of Economic Surveys, 1999assets.press.princeton.edu