A touted inherent characteristic of most public blockchains is their decentralized system of governance. Governance refers to the protocols which determine how the network accepts, processes, validates, shares, stores and secures information or transactions.
Traditionally, transactions (whether financial or non-financial) are carried out with the help or assistance of the state, banks and other organisations. These institutions have been fundamental to coordinating transactions and achieving consensus. Therefore, they usually control access to and the flow of information, as well as the processing and storage of information, resulting in centralized (hierarchical) system of governance where power is in the hands of a single or few organisations.
The financial crisis of 2008 shed much light on the behaviour of these institutions. Unsurprisingly, many blockchain enthusiasts and proponents view these organisations as inefficient, unnecessary, corrupt, abusing power, benefiting a few people, etc, and suggest that transactions can be carried out without the authority of such institutions. They propose a horizontal system of governance which is achieved through decentralization.
In politics, decentralization refers to the devolution of power from a central authority to a subordinate one. The power transferred may relate to decision making, taxation and revenue generation, administrative responsibilities and accountability. Governments may decentralize for a number of reasons including making governance more participatory and inclusive; to generate more revenue and to make financial management more accountable.
In computing, decentralization is made possible through peer to peer (P2P) networking. Unlike a centralized computing system where all nodes are connected to one server which performs the functions of processing, managing and storing data, P2P networking allows these functions to be distributed among all nodes.
The Blockchain technology makes use of P2P networking. Many erroneously assume decentralization of blockchains is simply achieved with the elimination of middlemen (e.g. the state and banks) and sharing of copies of the ledger with all participants on the network. Decentralization goes beyond this and involves the (even) distribution among all participants or users of the blockchain, the functions of accepting, processing, validating, viewing, storing and securing information or transactions and the power to make and enforce rules.
Vitalik Buterin, the founder of Ethereum, in ‘The Meaning of Decentralization’ identifies three types of decentralization: Architectural, Political and Logical.
- Architectural decentralization refers to the number of independent nodes (physical computers) on the system and how the system can withstand a failure of these nodes. Intuitively, more nodes are preferred to less but this does not necessarily give rise to decentralization.
- Political decentralization is the number of people or organisations who control the nodes on the system. Essentially, a variety of people or organisations in charge of the nodes is preferred since that would make it difficult for any one group to hijack the rule making functions of the system.
- Logical decentralization refers to interactions and data structures presented by the blockchain. Buterin notes that the chain should present a ‘single monolith’ and blockchains are logically centralised since ‘there is one commonly agreed state’ of the ledger.
Srinivasan and Lee (2017) go further to quantify decentralization. They identify the key aspects or subsystems of the blockchain (Figure 1) in which functions and activities must be decentralized and develop the Nakamoto coefficient as a measure of decentralization. The subsystems are the different interrelated groups which must work together to achieve the objectives of the blockchain. In their opinion, if any of these subsystems is not decentralized, the entire blockchain is centralized.
Figure 1: Subsystems of a Public Blockchain
Source: Srinivasan and Lee (2017)
The discussion below of what these subsystems are is not limited to Srinivasan and Lee (2017).
Client: This refers to the software (code base) that nodes use to send, receive and validate transactions on the blockchain. There are official and unofficial code bases. An official codebase is an original software developed and implemented by the developers of a blockchain while unofficial ones are developed and run by a third-party. Ethereum has three official codebases; Geth, Cpp Ethereum and PyEthereum while Parity is an unofficial codebase. Bitcoin’s official codebases are Bitcoin Core (which is necessary for a device to act as a full node) and Bitcoin Unlimited while unofficial ones include btcd and bcoin. Having more than one code base may be an advantage, especially if there are significant problems in one which can affect the entire blockchain. Srinivasan and Lee (2017) noted that both Bitcoin and Ethereum are centralized with respect to the client subsystem; 76% of Ethereum clients run Geth and 16% Parity.
Developers: They are the people that commit to developing and maintaining the software used in running a blockchain. Azouvi, Maller and Mekklejohn (2018) found that 7% of the Bitcoin Core software was written by one developer compared with 20% of Ethereum’s meaning that in writing and maintaining the software, the Bitcoin Core software was more decentralized than Ethereum. Buterin (2017) also suggests that developers of a blockchain should not be from the same organization and if possible, most should be volunteers. He goes on to say that rules about the fundamental or basic properties that the blockchains protocols should have must be clearly delineated. Developers have much control over the blockchain in its initial stages. They determine among others, the size of blocks and size of rewards of mining. Nonetheless, they must not have 100% control over the system. Lockyer (2018) notes that blockchain protocols must be written in such a way that the power of developers declines over time. Lastly, changes in the protocols of the blockchain (e.g. the consensus algorithm) must be approved by a majority consensus by users.
Ownership: This is measured by addresses on the network. Srinivasan and Lee (2017) chose a threshold of accounts with exactly 185 Bitcoins or more and exactly 2477 or more Ether. Dramaliev (2017) notes that ownership can influence the ability to manipulate the price of a cryptocurrency. Over 80% of addresses own less than 1 BTC i.e. 4% of distributed coins (Available at https://bitinfocharts.com/top-100-richest-bitcoin-addresses.html).
Exchanges: These are platforms which buy and sell (trade) cryptocurrencies. Srinivasan and Lee (2017) use the volume of a particular cryptocurrency traded in a particular exchange, in a 24-hour period as a measure of decentralization of exchanges.
Nodes: Srinivasan and Lee (2017) suggest the number of nodes connected to the system should be evenly distributed geographically. About 51% of nodes connected to the Bitcoin blockchain were in three countries; US, Germany and China (24.32% – US, 17.76%- Germany and 8.98%- China as at 2:44pm on 3/07/18 according to https://bitnodes.earn.com/). A recent study by Gencer et al. (2018) showed that Ethereum’s nodes are better geographically dispersed compared to Bitcoin’s. Also, 56% of Bitcoin nodes are in data centres compared with 28% for Ethereum, indicating that corporate organizations may be the owners of these nodes. Furthermore, it is important to have a good amount of independent full nodes i.e. computers which have the full and entire ledger of the blockchain.
Full nodes do not deviate from the chain’s consensus rules and are responsible for checking that the rules have been correctly followed during transactions. For example, they verify that transactions have the correct digital signatures, transmit new transactions to miners and ensure miners have followed the right rules to find the nonce. Any transaction which does not meet the consensus rules is rejected, invalidating an entire block on the chain. Miners can also run full nodes hence the need to have a viable number of independent nodes to act as a check on miners.
Miners: They solve a difficult computational problem to validate transactions on the blockchain and are rewarded with cryptocurrencies or transaction fees.
The advantages of decentralization include security of the leger since it exists in many different nodes, less vulnerability to a single point of attack, and a fairer system of allocating rewards. Recently, the decentralized nature of blockchains has been questioned, particularly with respect to mining. Although it is said that anyone on the network (even if using a home computer), can mine cryptocurrencies, it is increasingly clear that mining is dominated by just a few companies. Mining is a capital-intensive activity requiring specialized computer chips and an uninterrupted supply of electricity. Unlike small miners, these companies design and make their own chips used in computers for mining. Bitmain and Bitfury are two of the largest Bitcoin miners. BTC.com and AntPool (mining pools owned by Bitmain) controlled almost 40% of Bitcoin’s hashrate (mining power) as at 14 June 2018 (Figure 2). Three (3) miners control over 60% of Ethereum’s hashrate according to Gencer et. al (2018).
Figure 2: Share of mining pools in Bitcoin mining as at 14 June 2018
The dominance of a few companies gives reason to worry about the integrity (security) of the blockchain. Anyone with control of 51% of the chain’s hash power can change and distort information on the chain. What prevents these companies from colluding?
To answer the question of whether decentralization of public blockchains is a truth or an illusion, I am of the opinion that decentralization is multidimensional and achievable to an extent as the technology is still in its infancy. We should compare the use case of a blockchain with its subsystems. If a particular subsystem is not pertinent to the use case of the chain, then it’s centralization should not be the main reason for saying the entire chain is centralized. However, since one of blockchain’s touted advantage over existing ledgers is decentralization, there is the need to ensure that all subsystems of the blockchain meet the correct definition of decentralization over time.
Azouvi, S., Maller, M., and Meiklejohn, S., 2018. ‘Egalitarian Society or Benevolent Dictatorship: The State of Cryptocurrency Governance.’ University College London. Available at https://fc18.ifca.ai/bitcoin/papers/bitcoin18-final13.pdf
Buterin, V., 2017. ‘The Meaning of Decentralization.’ Published on 6 February 2017 on Medium.com. Accessed 25 June 2018. Available at https://www.google.com/url?q=https%3A%2F%2Fmedium.com%2F%40VitalikButerin%2Fthe-meaning-of-decentralization-a0c92b76a274
Dramaliev, V., 2017. ‘Building a Truly Decentralized Blockchain.’ Published on 2 October, 2017 by Medium.com. Available at https://blog.aeternity.com/building-a-truly-decentralized-blockchain-47ed6bd1f166
Gencer, A.E., Basu, S., Eyal, I., van Renesse R., and Sirer, E. G., 2018. ‘Decentralization in Blockchain and Ethereum.’ Cornell University Library. Available at https://arxiv.org/pdf/1801.03998.pdf
Lockyer, M., 2018. ‘The Decentralization of Value Through Cryptography and Blockchain Technology.’ Published on 6 January 2018 on Medium.com. Available at https://medium.com/@mattdlockyer/the-decentralization-of-value-through-cryptography-and-blockchain-technology-1c56e7c77b5e
MilkenInstitute. 2018. Cryptocurrencies: Irrational Exuberance or Brave New World? [Video File]. Available at https://www.youtube.com/watch?v=MMZGQSWD4tU [Accessed on 7 June 2018]
Srinivasan, B.S., and Lee, L., 2017. ‘Quantifying Decentralization.’ Published on 28 July 2017 by 21 Inc. Blog at Medium. Accessed 2 July 2018. Available at https://www.google.com/url?q=https%3A%2F%2Fnews.21.co%2Fquantifying-decentralization-e39db233c28e
The contents of CryptonomicsAfrica.com provide general as well as detailed information on cryptocurrency and blockchain technology. This should not in any way be treated as investment, business or legal advice. Users are encouraged to verify the information provided before acting on same. CryptonomicsAfrica.com would not be liable for any loss to any individual or entity occasioned by a reliance on the information provided as same is done solely at the risk of the individual or entity in question.