A Brief Discussion on the Mechanism Design of Bitcoin
By: Chang Jia (8btc.com)
Translated by: James Choi (8btc.com)
Edited By: Elizabeth T. Ploshay (Bitcoin Magazine)
The fierce “arms race” of the Bitcoin network has triggered concern over network fairness. At the beginning of Bitcoin’s birth, every individual could mine Bitcoin easily with a personal computer. According to Hal Finney, an early partner working with Satoshi, he gained thousands of Bitcoins within a few weeks by mining with his personal computer. Yet, today Bitcoin has become the “private mine” of mega mining corporation and professional miners who own many ASIC devices, which to some appears unfair. People also worry that Bitcoins are concentrated in the hands of a small group of people, which in turn hinders promotion of the currency. The Coase theorem states, however, that with zero or little transaction cost and clear property rights, the market could reach its Pareto efficiency of resources allocation, no matter to whom the property is given. A system would also be fair if the ownership of the rights of property is established from the start.
Satoshi did thoroughly consider the design of the Bitcoin mining mechanism. He was confronted with two options: one allocating Bitcoins to all users according to the number of the nodes, namely one-IP-address-one-vote; the other allocating Bitcoins to the miners according to the computing power, namely one-CPU-one-vote. He chose the latter considering the security of Bitcoin network. He wrote in the White Paper that “If the majority were based on one-IP-address-one-vote, it could be subverted by anyone able to allocate many IPs.” Miners obtain Bitcoins by processing transaction data, which may be viewed as an incentive mechanism. In retrospect, however, miners should put in their computing power to obtain Bitcoins which could be deemed as a public resource. Since the security of the Bitcoin network is proportionate to the network’s computing power, the computing power put in by miners are used to ensure the security of the network, enabling normal users to trade safely without putting in any computing power. This is fair to both the users and the miners.
Although the initial allocation of Bitcoins was established by the protocol, the market decides final allocation. Users possess the right to collect computing power from miners and this right can be sold, meaning miners “pay” the computing power to users in exchange for the ownership of Bitcoins in the block. On the other hand, miners possess the right to obtain Bitcoins in a block and this right can also be sold, which means users can then purchase Bitcoins from miners without putting in computing power. If the computing power paid by miners is higher than the revenue generated by mining, miners would put in less computing power. If the expense paid by users to purchase Bitcoin is higher than the mining cost, more and more users would join the mining industry. The market would reach equilibrium at the end and a highly-efficient allocation between computing power and Bitcoins could be reached.
In the first days after Bitcoin was created, transaction cost was almost zero and some special users could realize certain goals with small-amount-transaction. For example, Wikileaks coded some secret files in the blockchain with huge quantities of small-amount-transactions; gambling websites like SatoshiDice informed their users of the gambling results (or even sent commercials) with small-amount-transactions. These past actions have put more burdens on the blockchain. Some view that those users who overused the small-amount-transactions have taken too much public resource, and they support a transaction tax. This Pigouvian-Tax-like design can reduce the junk transaction to a huge extent and equalize the private cost and social cost of the transaction.
The rule of levying the transaction tax of Bitcoin, when calculated is economical. It would discourage trivial payment, meaning that if the output of a payment is less than 0.01BTC, a transaction fee of 0.0001BTC would be collected. A payment with Bitcoins having a longer history and value is prioritized. The trade data is then “weighed” and a fee is collected for each kilobyte. As a unit tax, the transaction tax could be levied relatively ideally with the adjustment of deft mathematical algorithm. Consequently those who do a lot of small-amount-transactions must refrain due to the high costs, while regular users can still keep the cost edge compared to traditional ways of payment. Thus, micro-adjustments can still be made by updating the protocol.