Centralization and decentralization of rewards

Note: this is not intended to be a comprehensive review of the ways markets centralize and decentralize, but rather a list of some of the ways they do centralize, and a remark on how decentralization at a given level often results in centralization at another level.

Centralization

There are three reasons why rewards tend to centralize in the hands of few:

• Popularity used as a proxy for trust: when people have to choose between service/product supplier/distributor they do not know, and do not know if they can trust, they tend to use popularity as a proxy for trust. Because people choose the most popular provider, the most popular becomes even more popular, attracting even more customers, and so on. This takes place when trustworthiness is difficult to evaluate without a previous interaction.
• Value depends on usage: for some products, their value depends on the number of the people using them. For example, the more people use a fax machine, the more the fax machine is worth. This takes place when the value of a product comes from users interacting with each other or when the product is a tool used by a new class of professionals to produce or distribute another product (e.g. an App Store).
• Economies of scale: some products or services have a high fixed cost and a low variable one. In such case, selling a lot of products allow the fixed costs to be spread amongst them, allowing the producer to charge users a lower cost, which allows it to sell even more product, etc. This takes place when products have a high cost component which is fixed.
• Compounding: resources accumulated at a previous stage can be invested to reap more returns, which can be reinvested, and so on. This does not only apply to money, but to knowledge, networks of people, etc.

Decentralization

It follows that some events that might lead to a decentralization could be:

• The raise of aggregators as trust providers: if aggregators, such as Uber or AirBnB, can provide users a way to trust service providers (e.g. drivers and landlords), then users do not need to use popularity as a proxy for trust. In this case, the suppliers’ market flattens, whereas centralization of rewards is shifted to the aggregators’ markets (landlords or car-for-hire markets become flat, but AirBnB and Uber get most of the profits at the upper layer).
• Small batch production: the raise of suppliers who can produce different products or services at small batches without incurring into big costs will produce a linearization of the rewards in the markets where the suppliers were specialized in big batches for niche markets, whereas the upper layer gets centralized.
• Staggered tax brackets: can offset some of the effects of compounding.

It is worth noting that often, solutions to centralization at a given level do not destroy centralization, but just shift it to another level.