Marginal contribution nonsense


Consider two imaginary artists, Jay and Ani. Jay raps about problems not created by women and Ani sings ballads about problems created by men.

Unequivocally Jay is more popular. At almost any time and place more people want to listen to Jay than Ani. Of course popularity is not the only measure of worth. We could make qualitative arguments as to why Ani’s work should be more prominent, and these ultimately translate to predictions about hidden preferences: Maybe everyone will value her more in the future, or some people value her really strongly, or would value her if it weren’t for marketing. We’ll return to these concepts later.  But for now let us concede that the market has ranked Jay’s contribution higher than Ani’s.

How much higher is Jay’s contribution relative to Ani’s? Today, Jay is 50 times richer. Does this mean his contribution to the world through music is 50 times greater? Not so fast!

Suppose our wonderful artists stated their careers in the 1900’s. There’s no amplification and no recording, so all they can do is acoustic concerts in relatively small halls. Jay is still more popular and manages to book 3 times as many gigs. In these circumstances where the artist’s product scales with their labor we can say that yes, Jay’s contribution is 3 times Ani’s or close enough.

Then it’s 1950 and we have amplification. Now people can gather to really big concerts in stadiums. Jay can pull in the big crowd and earns 10 times as much as Ani who’s still doing small gigs. Did Jay just expand his contribution, or did the work of engineers and athletes and builders do that?

Fast forward to 1980 and someone invents CDs. Imagine at first there’s only two CDs on the market, Jay’s work and Ani’s work, and they’re very expensive. Let’s say they’re $100 so consumers can only afford one. Forced to choose, the vast majority of people go for Jay’s CD and he ends up selling 100 times as much as Ani. Is his contribution now 100 times greater? Why? What changed since the acoustic days?

A few years later in the 1990s publishers have figured out how to make CDs cheaply and price them properly, so now they’re $10. Suddenly consumers can afford to buy both Jay’s and Ani’s work, and Jay is now making 50 times Ani’s sales. Did their relative contribution actually change?

Now it’s 2010 and some geeks invented streaming music. They charge $10 a month and let you listen to whatever you like, distributing money to artists according to how often each song is played. Now people have access to both Jay’s and Ani’s work and it turns out that Jay’s work gets played 200 times more than Ani’s. Wait, what happened? Jay now earns 200 times as much as Ani. Does that reflect his contribution, or is his music more everyday entertainment compared to Ani’s deep stuff?

Now let’s revisit the qualitative questions about worth that we parked at the start. Is it correct to reward artists according to how many times their songs get played? What if people value listening to Ani’s songs occasionally, or just having the option to listen? Remember people don’t pay per listen, they pay a flat fee. What if artists were rewarded progressively, say by the log or square root of play counts? What if the money was apportioned by listener ratings, or an even more direct listener choice to support specific artists? Changing the model of what’s essentially interpretation of the subscriber’s preferences completely changes the valuation of each artist’s work.

The moral for economists is that marginal contribution is nonsense. A meritocratic market can rank contributions with plausible veracity but says nothing about the magnitude of contributions. Earnings are an emergent result of technology, ownership, legal structures, trust, marketing, and other factors. These factors determine the shape of the earnings curve – how much winners win and losers lose. A meritocracy, at best, defines who the winners are.