The New Yorker's Mary Norris went to Padua for an auction run by Coys of Kensington, auctioneers of vintage motorcars. She accompanied Giuseppe Favia, who was selling a 1959 Mercedes-Benz 190 D with literally sainted provenance. The auction itself, however, was not quite as blessed.

Bidding began at a hundred and fifty thousand euros, with the auctioneer under pressure to push the price up. (It had been widely reported that a Volkswagen Golf said to have been owned by the Pope had sold on eBay for a hundred and eighty-nine thousand euros.) By the end, three bidders were contending, one of them on the phone from Belgium. The final bid of two hundred and forty thousand euros seemed to come from a bidder in the hall. When it was all over, Signor Favia tried to find the bidder, but he was not in the crowd.
Coys not only protected the anonymity of the buyer but maintained, despite the throng of witnesses, that the car had not been sold. David Barzilay, of Coys, invoked a technicality: “The auctioneer never said ‘Sold.’ ” When pressed, he added that the reserve of three hundred thousand euros had not been met. Apparently, it is not unusual for negotiations to continue after an auction; the auctioneers had their commission to consider, and they might not have heard the last of the Belgian. This left Signor Favia in a state of limbo.

This is more than a little confusing, but it does give me the opportunity to clear something up which I've been meaning to get to for a while. I talked a bit about reserve prices and the pratice of "taking bids off the chandelier" in my guide to deciphering auction results earlier this month. But (as ever) there's more to be said – and the key point is that if an item ever fails to sell at auction, the final bid is always fictional. Is that a bad thing? Daniel Grant would have you believe it is, when he complains about "phantom (or 'chandelier') bids that the auctioneer announces and records to get the bidding up to the reserve". But in fact it's not a bad thing at all: in fact, it's a necessary thing if the auction, as designed, is to work..

Remember that an auction, like any market, is an exercise in matching buyers with sellers. The buyers, naturally, have a price above which they will not buy; the seller, equally naturally, has a price below which he will not sell. If those two prices don't overlap, there's no deal: the item can't be sold. If they do overlap, then the auction mechanism is designed to discover which buyer is willing to pay the most money.

That's done with the bidding system: each would-be buyer outbids the last, until there's only one left. The system works well, when there are multiple buyers. But it doesn't work very well when there's only one would-be buyer, even if he's willing to pay more than the reserve, because there is no one else to bid him up to the reserve level at which a deal can be struck.

That's where the auctioneer comes in. He tells everybody, at the beginning of the auction, that he "may continue to bid on behalf of the seller up to the amount of the reserve either by placing consecutive bids or by placing bids in response to other bidders". That way, even if there's only one real bidder, the auctioneer can take bids off the chandelier up to the reserve price and keep that one bidder bidding. The purpose of this is exactly what Daniel Grant complains that it is: it's to get the bidding up to the reserve price. But remember, that's exactly what both the buyer and the seller want: they both want a deal acceptable to them. If there were no phantom bids, then the item for sale would simply not get sold.

Even if there are three real bidders, if only one of them is willing to go over the reserve price, then it might well be necessary to invent a chandelier bid or two in order to discover that fact. And there's simply no way that an auctioneer will let an item to "pass" unsold without bidding one more time, on behalf of the seller, to see whether that final bidder might be willing to go a bit higher and get the deal done. There is no harm done when that happens: if the deal does get done then everybody's happy, and if it doesn't then that's the same outcome as if no phantom bid had been made.

In Padua, then, the bidding reached some level below the reserve level of €300,000, and the auctioneer put in a final chandelier bid of €240,000 to see if the mysterious Belgian – or someone else – might be willing to go higher. When that was not the case, the vehicle was passed – something which auctioneers normally say quickly and quietly, since they don't like to admit that the sale didn't get done.

If Signor Favia was trying to find the bidder, then, it wasn't because he thought he had just sold his car for €240,000 – he knew full well what the reserve was; indeed, he certainly set that reserve price himself. The auctioneers do indeed have their commission to consider – which is why, given their druthers, they'd more than happily sell the car for €240,000 and pocket that certain commission. Norris's implication – that the auctioneers could have sold the car but didn't because they wanted a bigger commission – simply makes no sense. People who regularly attend auctions soon begin to understand all of this intutitively. But to the rest of us – including, it would seem, the New Yorker's fact-checkers – it can all get a bit confusing.