The logic of Marshallian S&D curves are wonderful in several respects:
- resolves the “diamonds and water paradox” (why does unnecessary jewelry cost more than necessary water?)
- sounds reasonable across a variety of real-world scenarios (FCOJ futures, corporate bond issuance, grocery stores, machine parts, olive oil exports, Tyler Cowen’s umbrella term “markets in everything”)
- actually works in experiments! The legend is that Vernon Smith used to say in class that Marshallian S&D was “just a theory” — and then was shocked that prices actually converged to the predicted
P*
Here’s a great way to misapply the Marshallian logic and arouse my ire:
- Say “Markets make sure that people who want things more are the ones who get them.”
That’s not what the theory says. We use the jargon willingness to pay or reserve price to talk theoretically about the maximum someone would counterfactually give up for something—and equate this (by rational consistency hypotheses) to how much utility they derive from obtaining it. (The experiments I mentioned above literally created a reserve price—a redeemable coupon for $13 if you get the paper at P*, so we as non-omniscient lab-gods know that you actually assign a personal dollar value on the good—and know what it is. So the fact that those experiments worked doesn’t prove the extra assumptions about the way people’s consent, pleasure, engagement, and desires interface with an opportunity for economic exchange.) Laura’s measured willingness to pay does not say how badly she wants something relative to Gemma. Why? Because maybe Gemma is poor and Laura is rich.
In the real world, rich people engage in retail therapy at prices that would pay for a poor person’s housing and food for months.
Maybe it makes them feel good, or they do it as a way to socialise (if you don’t consider yourself rich but you’re reading this on a computer: do you socialise at bars or restaurants or just outside on the street? Why?), or maybe they’re bored. Whatever.
Clearly we can’t give Gemma £100 and give Laura £100,000 and conclude that Laura wanted the dress more because she paid more for it. It might be reasonable if both were in the same place with the same financial resources.
The mathematics behind the S&D graph aren’t that complicated. (It does require thought—but not years’ worth of thought—to understand the Marshallian model.) But still, I think because of the transition from English → maths → English, and the jargon words interposed with normal words, the overall rhetorical effect is to cover the obvious fact of inequality whilst redirecting attention to “optimal” (another jargon word budging in on the default namespace!) allocation.
The hypothesis of logarithmic utility per individual has been around since the 1700’s at least. (Implying €1000 means more to a poor person than to a rich person.) And yet people still use this fallacious reasoning that markets allocate goods to those who “want it the most”.
Sorry: willingness to pay is a function of both desire and of ability to pay.
