I've been thinking through an idea for a few years now, and am about to put out some writing on — I'd love to get some feedback.

In a nutshell, it's market-based transport load balancing mechanism with dynamic, agent-specific pricing.

👇

For context, I'm really thinking about physical transport networks here — urban road networks. That said, I think a lot of these ideas would apply to other types of transport networks (like the Internet, maybe).

Questions:
How important are decisions about the route taken between an origin and destination to network load measurements?

How would the introduction of a route-specific, time-dependent cost affect agent behavior? 

Could precision congestion pricing modify agent incentives and therefore improve measures of network operational efficiency?

Right now, I'm trying to wrap my mind around the nature of roads as public goods. It's one of the most commonly-cited examples:

Right now, I'm trying to wrap my mind around the nature of roads as public goods. It's one of the most commonly-cited examples:

https://www.investopedia.com/terms/p/public-good.asp

But this misses something: at a certain level of usage, a new user *does* reduce the availability of roads to others. 

Put another way, an area of road, like any physical space, is rivalrous. In certain contexts, "its consumption by one consumer prevents simultaneous consumption by other consumers".

But this misses something: at a certain level of usage, a new user *does* reduce the availability of roads to others. 

Put another way, an area of road, like any physical space, is rivalrous. In certain contexts, "its consumption by one consumer prevents simultaneous consumption by other consumers".

https://en.wikipedia.org/wiki/Rivalry_(economics)

A market-based load balancing mechanism is most useful in network states further up the rivalry continuum, i.e. as load increases and congestion forms.

The idea is hatched from my past (and current!) work on location proofs, some experiments with network routing algorithms, and the insight that smart contracts offer both programmable identity (accounts) and a space to implement market makers that employ sophisticated pricing models (bonding curves).

I'll share more as the thinking develops, but I would love anyone to ask questions, point out flaws, sharpen my theoretical thinking etc 🙏

My hypothesis is that a well-designed mechanism would allow for the precision management of network loads based on a range of factors, and that these advanced network access pricing systems could improve network efficiency and equity.