Update: “Lord Keynes” provides a great explication of Kaldor’s theoretical work on this subject.
Back in the day when I was running a high-tech conference company, we had a favorite (and actually rather cruel) interview question:
“What’s the best price for a conference?”
There was only one right answer:
“The price that makes us the most money.”
That answer encapsulates the position of almost every business trying to sell real goods and services. You have to choose a price, and you have little or no idea what sales differences will result from different choices. The implications are enormous — no other decision affects profits so powerfully — and you’re basically shooting in the dark.
Every business or product-launch plan I’ve ever seen has a huge dartboard right in the middle of it: How much, how many, will we sell? (“Well, it depends what we charge…”) In some businesses there are ways to do controlled tests of different prices and see how sales respond — Amazon being a brilliant example, and we did a a bit of it using direct mail with split runs — but most businesses (i.e. all of Amazon’s producers/suppliers) don’t have that luxury. You have to just choose a price with your best guess, based on various scraps of hard-to-interpret historical-sales and market data. It’s incredibly frustrating.
Which brings me to the point of this post: most producers are dealing (at least in the short term whose length varies with the type of business) with a fixed-price market. Once they’ve set their price, they can’t go changing it all over the place.
So: They’re not receiving any of the market’s supposedly informational price signals. They’re receiving quantity signals. That’s the information that producers derive from the market. Then maybe they change the price, and get more quantity signals. But again, those signals are always hard to interpret because you don’t generally have a controlled test to know whether the price is what drove quantity changes, or whether it was something(s) completely other.
Price is fixed, while quantity is very flexible. i.e. Analysts expected Apple to sell five million iPhone 5s in its first weekend. They sold nine million. Did the price go up? No. They rousted workers out of bed in China and filled the goddamn orders. When one of our conferences wasn’t selling well we couldn’t just lower the price, cause we’d piss off everyone who’d already signed up. If sales were good and it looked like we might sell out (there was simply no more room for hotel employees to place chairs), the last thing we were going to do was raise the price and risk stomping on that success. It’s very difficult for producers to derive prices signals from the market.
This is utterly unlike the market for financial assets, where price is infinitely and instantaneously flexible, while quantity — i.e. the number of Apple shares outstanding — is pretty much fixed and unchanging. (When you buy my Apple shares, the quantity or supply of saleable Apple shares is unchanged.)
In the market for financial assets, the price signal is (almost) everything. In the market for real goods and services, the quantity signal is (almost) everything.
There are lots of places to go with this thinking, but I’ll leave that to my gentle readers for the moment.
Thanks to Mike Sankowski for prompting this post.
Cross-posted at Asymptosis.