Tradeoff Between Decentralization and Scale Puts Bitcoin to the Test

Note: As this brief went to press, the SegWit2x proposal cleared a first hurdle, as a majority of Bitcoin miners signaled their support.

With the meteoric rise of Bitcoin over the past few years, the seams of the system have begun to strain. This growth has come to a head with extreme congestion on the blockchain. Not only has the waiting time for transaction verification increased, but the costs of these transactions have also increased. Users of Bitcoin are forced to offer increasingly large sums as transaction fees to catch the eyes of miners and buy their transactions a spot on the next block. Because the number of new bitcoins paid to miners as a reward for confirming transactions is programmed to decrease over time, transaction fees will grow in importance as compensation to miners. These rising fees are likely to be prohibitive for some prospective smaller users.

We have previously written about whether Bitcoin’s radically decentralized governance structure is equipped to implement changes to its code that would address this scaling problem. But as the debate rages on, a deeper potential problem has emerged: it may not be possible to significantly scale the network without adding a significant amount of centralization. As we discuss below, many Bitcoin stakeholders have derided the Emergent Consensus proposal, which would raise the limit on block size but possibly make mining much more expensive, leading to more centralization. But now, it appears that the major competing proposal, called Segregated Witness, would end up centralizing the governance structure in a different way.

Taming the Beast

Over the past year or so, Bitcoin developers have been discussing how to address the growing congestion and how to scale the Bitcoin system to accommodate its rising popularity. Bitcoin, according to a white paper by Bitcoin programmers Joseph Poon and Thaddeus Dryja, can typically handle no more than 7 transactions a second, while large banking systems such as Visa have been rated to handle 47,000 transactions every second. If Bitcoin or any other cryptocurrency hopes to become widely adopted, it must be able to compete on the same level as today’s leading payment systems. The issue of scaling is therefore a critical concern in the Bitcoin community.

The two main proposals to upscale Bitcoin are known as Segregated Witness (SegWit) and Emergent Consensus. SegWit, put forward by a core group of Bitcoin developers, streamlines the information recorded on the blockchain to allow more transactions to fit within the current 1 megabyte size limit for a block. Emergent Consensus gives mining pools the ability to decide as a community whether to raise the 1 megabyte limit on block size. Support for Emergent Consensus was high at first because of its unlimited ability to scale Bitcoin. However, concerns over the politicization and centralizing of power in the network shifted support to SegWit.

In 2015, Bitcoin developers released a compromise called SegWit2x, which doubles the size limit of blocks to 2 megabytes while maintaining the streamlining features of the original proposal. The SegWit code is planned to be released into the Bitcoin network on July 21. If the implementation goes off without a hitch, the 2 megabyte block-limit increase involved in SegWit2x will occur around November 10.

Ride the Lightning

The SegWit2x proposal introduces the Lightning Network, a proposal expected to allow massive scaling for Bitcoin and facilitate cheaper and smaller transactions between millions of users without clogging the Bitcoin system. The proposal is to allow these micro-transactions to occur off the central ledger, with only the net movement of money recorded and verified on the blockchain. The hope is that this would allow small transactions without prohibitively high transaction fees.

The proposed LN works by having interested partners open transaction accounts where they exchange funds without reporting to the blockchain until they decide to close the account. The opening of the account is recorded on the blockchain, and the parties involved set aside the maximum amount of money that can be used in the transactions. The two parties keep time-stamped receipts of each new transaction, which they can redeem at any time on the blockchain. Once the receipt is redeemed, the money held in the account is divided between the involved parties according to the balance recorded on the receipt. The LN has multiple features in place to ensure that one side does not fraudulently redeem an older receipt that does not reflect all payments.

The Lightning Network shows promise in allowing individuals to engage in nearly costless micro-transactions, but it would not be efficient for people to open accounts for one-time transactions. Rather than requiring users to create new accounts every time they wish to transact with someone new, the LN allows the same transaction to be made through multiple accounts. This allows for the existence of an intermediary that would have accounts with consumers on one side and retailers on the other. Small retail transactions that don’t clog the network or require high transaction fees would then be possible.

More Centralization

But while the Lightning Network solves some problems in Bitcoin’s architecture, it creates others. First, some have questioned how scalable this system really is. It is only possible to transfer as many funds as are initially put in the account when it is opened, making it advisable to overestimate the funds one will need. Unless it becomes possible to open accounts using credit, it does not seem feasible for the ordinary user to have more than a few accounts open at a time. In addition, if Bitcoin truly became a global currency and truly billions of people were exposed to this system, the blockchain likely couldn’t handle more than a few connections per person. Allowing only a handful of accounts for each person makes it difficult to connect a large number of people without the development of centralized “banks” through which thousands of transactions are sent between individuals. This is not necessarily a bad thing, but it flies in the face of the decentralized model of Bitcoin that hardcore users and miners currently support. (Some of the math to back this up can be found here.)

The potential problems go beyond centralization. If billions of people across the world adopted Bitcoin and the Lightning Network, the necessary centralized “banks” would need extremely large sums of money on hand to create accounts with individuals and retailers. In addition, holders of these accounts would have to wait until they were closed to access their funds. One possible solution might be if these intermediaries functioned more similarly to credit card providers, though this would involve further (perhaps contentious) changes to the code and additional layers of complexity that could impede consumers and firms from adopting the bank-like solution.

No Perfect Solutions

The Emergent Consensus solution would require higher capital costs to effectively mine bitcoins. The Lightning Network (SegWit2x) would require bank-like intermediaries for the system to function at a truly global level. One of these proposals may be the right scaling solution for Bitcoin, in which case the currency’s evangelists must partially let go of the initial vision of radical decentralization. Importantly, under either proposal Bitcoin would remain far more decentralized than current currencies or payment systems, as the distributed nature of the database would remain intact.

A third possibility is that solutions down the road, for Bitcoin or other cryptocurrencies, will emerge to solve the scaling problem without significantly centralizing the system. This is not a make-or-break moment for blockchain-based digital currencies, but it is an important crossroads. The collision of Bitcoin’s radical decentralization with the reality of large-scale use has spurred and will continue to spur important innovation.

Related Articles – Cryptocurrency