A Decentralized Fair Governance Model for Permissionless Blockchain Systems

Abstract

Blockchain systems are a new way to reduce or even eliminate the role of the middlemen in an eco-system. For example, the Bitcoin, as one of the most well-known blockchain platforms, shows that it is possible to transfer money without the need of any (intermediate) bank at all. More generally, it allows for the decentralization of roles. In this paper, we focus on permissionless blockchains, which are systems that allow participation without upfront approval on other parties, as opposed to permissioned blockchains. Permissionless blockchain systems support direct business transactions between peers, so without any intermediate and centralized entity very well. However, the organization of the governance of such systems is less obvious. We argue that, in order to arrive at a really decentralized eco-system where power is fairly distributed, the governance should be decentralized; in other words, it should not be in the hand of one controlling entity. In this paper, we analyse, in a model-based way, for three well-known blockchain systems (Bitcoin, Ethereum, Tezos) the governance processes. Based on this analysis, we draft an improved governance process for permissionless blockchains.

-

Networks of enterprises and end-users can be viewed as eco-systems : \collection of companies that work cooperatively and competitively to satisfy customer needs" [ 1 ]. Many of such eco-systems are fairly centralized, and often called platforms. In [ 2 ], a platform is considered as \a set of components of a bigger system". We consider a platform as an eco-system itself that provides services to some other eco-system. Examples of platforms include Google/Android, Apple/IOS, Facebook, LinkedIn, Uber, AirB&B, and many more.

All these examples are controlled by a single enterprise, although they have many participants. We call such platforms centralized platforms as they are governed by one party only. Centrally led platforms can easily lead to symptoms of value extraction: Generating income not by producing anything new but by charging above the competitive price, usually by exploiting a monopoly [ 3 ]. Although companies such as Uber and AirB&B do add something new (namely a matching service), they do not own taxis or hotels themselves. They achieved a very large market share, often by initially o ering services for a very low price of even for free, to increase prices signi cantly, once a dominant market share is obtained. This is symptom of value extraction.

It is doubtful whether value extraction behavior of many platforms is benecial to society. At least, it would be useful to have facilities for enabling decentralized platforms and eco-systems such that decision power is fairly distributed over the participants. This would make it much more di cult to expose value extracting behavior and can serve as aid for eco-systems that want to o er a fair alternative to centrally governed eco-systems.

Blockchain is a technology that allows to build peer-to-peer eco-systems, that is: without intermediate parties, where participants do not need to trust each other on beforehand but instead rely on crypto-techniques to create trust. The most well-known example is the Bitcoin [ 4 ] but many alternatives exists. Often, a distinction is made between permissioned and non-permissioned blockchains. For the latter, everyone can participate, whereas for permissioned blockchains, others have to approve participation of a user on beforehand. By o ering peerto-peer business transactions without the need of an intermediate party such as a bank, blockchain can be considered as a strongly decentralized technology, enabling also a fully decentralized eco-system.

To support decentralized eco-systems to the maximum extent, there is however more functionality needed than just peer-to-peer transactions without an intermediate party. In particular, the governance of the operational peer-topeer system should also be decentralized. We de ne governance in decentralized eco-systems as \the set of rules a system has to obey, and which are set by another system" [ 5 ]. Potentially, elements (e.g. participants) of both systems may overlap. Moreover, in case of decentralized eco-systems, governance has to be decentralized too, and preferably in a fair way, as often the justi cation for a decentralized eco-system is the avoidance of power-concentrations as often seen in platform-oriented centralized eco-systems.

Unfortunately, the governance of non-permissioned blockchains such as the Bitcoin and Ethereum is (1) not fully decentralized, and (2) quite informal. With respect to informality, for the Bitcoin there is the mechanism of Bitcoin Improvement Proposals (BIPs) and Ethereum has a similar mechanism. But still, these mechanisms lead many things open to the imagination.

The contribution of this paper is twofold: First, we formalize the governance processes of three non-permissioned blockchains, namely Bitcoin, Ethereum, and Tezos with the goal to analyze them with respect to fair governance. Second, we use this analysis to propose a fair governance process for non-permissioned blockchains.

Fair governance in eco-systems

We de ne governance in decentralized eco-systems as \the set of rules a system has to obey, and which are set by another system" [ 5 ]. Governance comes with setting rules on the hand, and continuously monitoring whether systems comply to the stated rules on the other hand. A system can be a person (e.g. a director that sets rules for an employee and on a monthly basis evaluates whether the employee satis ed the rules), an enterprise (e.g. a large company and its shareholders), or a government (e.g. a municipality sets rules for its inhabitants).

From a modelling point of view, many perspectives can be taken on governance. For example, governance can be seen as formal contract that states the rules explicitly, and can be modelled, e.g. with Symbolio [ 6 ]. Since governance, if done well, is a substantial e ort, a sound business case should be present for the governance task itself, which can be represented by e.g. an e3value model. But in many cases, governance takes a behavioural, process oriented point of view mainly. Many rules and regulations are put into operation by means of procedures which can be represented by process models. In this paper, we utilize the Business Process Model Notation (BPMN) 2.0 [ 7 ] for that purpose.

We are interested in decentralized eco-systems, with a fair distribution of governance power, as a realistic alternative for the big tech-platforms. We de ne fairness in eco-systems following [ 8, 9 ] as follows: { Participation. Fair governance requires active involvement in the decisionmaking process of all who are a ected and other interested parties. It includes all participants interacting through direct or representative democracy. Such a broad involvement requires that the a ected be well informed and organized in order to participate constructively. Furthermore, participants should be able to participate in an unconstrained and truthful manner. { Rule of law. All participants should be treated equally and fairly by obeying the law. Fair legal frameworks, with its underlying democratic principles put no participants above the rules of law. Without rules, anarchy will prevail. Anarchy or chaos are caused by governments acting beyond their scope of power, and participants neglecting the law. { E ectiveness and e ciency. Fair governance ful ls societal needs by incorporating e ectiveness while utilizing the available resources e ciently. E ective governance ensures that the di erent governance actors meet societal needs. Fully utilizing resources, without being wasted or underutilized, ensures e cient governance. { Transparency. Information on matters that a ect their participants and whom they may concern must be freely available and accessible. The decisionmaking process is performed in a manner which is clear for all by following rules and regulations. Transparency also includes that enough relevant information is provided and presented in easy to understand forms or media. { Responsiveness. A responsive fair governance structure reacts appropriately and within a reasonable time frame towards its participants. This responsiveness stimulates participants to take part in the governance process. { Consensus-oriented. Fair governance considers the di erent participants' viewpoints before decisions are made and carried out. Such governance is de ned as consensus-oriented because it achieves a broad community consensus. In order to reach this wide consensus, a rm mediation structure, without any bias towards its participants, should be in place. { Accountability. Accountability is de ned as responsibility or answerability for one's action. Decision-makers, whether internal or external, are responsible for those who are a ected by their actions or decisions. These decisionmakers are morally or legally bound to clarify and be answerable for the implications and selected actions made on behalf of the community. 3

Fairness of governance in three non-permissioned blockchain platforms

To understand fairness of decentralized blockchain eco-systems with respect to governance, we have analysed three non-permissioned blockchain platforms, namely Bitcoin and Ethereum (as these are the two most popular platforms), and Tezos, which is in contrast to the rst two platforms a blockchain platform with on-chain governance.

For all three platforms, we have developed BPMN process models for the governance processes. To the best of our knowledge, this is the rst attempt to formally describe the governance processes of these platforms. For Bitcoin, the governance process is partly described by the Bitcoin Improvement Proposal (BIP) process [ 10 ]; for Ethereum something exists as the Ethereum Improvement Proposal (EIP) [ 11 ]. However, the process descriptions are mainly textual, informal, and high level. The actual processes are much more complicated, and not explicitly articulated. For example, many issues are settled as informal discussions on Reddit. Therefore, we have not only looked into the BIPs and EIPs, but also harvasted more informal information regarding decision processes in Bitcoin and Ethereum.

Tezos is considered as a self-amending distributed ledger. Unlike Bitcoin and Ethereum, Tezos has the ability to perform meta-updates to its own code. Metaupdating of its own code is done by incorporating its stakeholders' consensus through an on-chain governance mechanism [ 12 ]. Tezos' governance process of proposing changes and features is achieved by following the Tezos Amendment proposal process (see e.g. [ 13 ]). We analyzed the Tezos Amendment process mainly by the o cial Tezos Medium [ 14 ] and the Tezos governance explorer (see https://forum.tezosagora.org/).

For Bitcoin, Ethereum, and Tezos we modelled the governance processes using BPMN. Due to lack of space, we can not present the models in this paper, but they can be consulted online [ 15 ]. We limit ourselves to an assessment of fairness of the governance process based on the BPMN models in Table 1.

Ethereum (EIP)

On-chain Tezos (Amendment) Fair nance teristics Participation

O -chain gover- Bitcoin (BIP) charac

All participants have an Not all participants are ac- Only bakers (miners) are active role in the gover- tively involved in the gov- admitted to taking an acnance process. The pro- ernance process (e.g. users tive role in the governance cess employs a mailing-list and miners). Four discus- process. The process emto inform and exchange sion platforms are utilized powers discussions using knowledge. There are no to exchange and organize messaging and blogs. Baklimitations on who can par- information about propos- ers are limited in particticipate in a BIP submis- als. Anyone may partici- ipation, depending on the sion. pate in submitting EIPs number of rolls which prewithout any limits. vents power centralization.

Rule of Law BIP outlines the gover- EIP describes the gover- The amendment process nance rules and framework nance rules and frame- formally describes the govinformally with o -chain work informally with o - ernance rules and framegovernance. The gover- chain governance. The gov- work with on-chain governance mechanism is not ernance mechanism is not nance. The only identi ed always clear or precisely always clear or precisely exception is regarding hard described (e.g. status described (e.g. the number forks that can occur, but change from nal to of editors responsible per are not formally de ned. obsolete). EIP).

E ectiveness The case that hard, soft, Even though hard forks are On-chain governance e ecand e ciency and user activated forks not a part of the EIP gover- tively meets participants' occur is evidence that the nance process, they do oc- needs, which results in BIP governance process cur. These hard forks re- reducing the dependency is not always e ective. sult in not e ectively meet- on hard forks to exist. These forks cause ine - ing the participants' needs. Without hard forks, there ciencies by enabling multi- The outcome in uences ef- is no fragmentation in ple blockchains to exist. ciency by fragmenting the the blockchain, which prosystem in multiple co- motes e cient use of the existing blockchains. resources.

Transparency BIP's process provides EIP's process o ers the Amendment process inforthe necessary information required information mation can be perceived through channels such as through multiple dis- freely by anyone. The email and GitHub, which cussion platforms. The Amendment information can be freely perceived. platforms can be accessed can be perceived on multiHowever, the barrier ex- by anyone wanting to ple platforms, which o er ists in being informal, participate. However, the the same information. technical, and not clearly numerous platforms cov- It is presented formally, presented. ering di erent information organized, and elaborated causes information frag- on technical as well as mentation. Other barriers non-technical topics. are informality, technical jargon, and not easy to digest.

Responsiveness Two time frames are iden- One time frame is identi- Five time frames are identi ed in BIP's process cov- ed in EIP's process cov- ti ed in the Amendment ering only two speci c de- ering only a speci c de- process encompassing the cisions in the process and cision in the process and whole governance process. not the whole governance not the whole governance These precise time frames process. The lack of time- process. The lack of time result in a responsive govframes for the remaining frames for the remaining ernance process. This reprocesses results in poor processes results in poor sponsive governance congovernance responsiveness. governance responsiveness. structively a ects particiThis poor responsiveness This poor responsiveness pation. negatively a ect participa- negatively a ect participation tion

A fair governance proccess for non-permissioned blockchain platform

An interesting question is if we can construct a better (meaning a more fair) governance process based on our analysis of the Bitcoin, Ethereum, and Tezos governance process. For this new governance process, a signi cant part is from Tezos as this (1) an on-chain process (and hence can be automatically supported and thus speci ed as such) and (2) this process is best articulated. The rest of the proposed process is a combination of Bitcoin's and Tezos' governance processes. The combinations are from the participation, rule of law, and consensus-oriented characteristics. Our proposal, represented as BPMN model can be found online [ 15 ]/ Below we summerize the highlights of our improvement proposal. { Participation. A combination of Bitcoin's and Tezos' governance process will be useful to ful l participation. Bitcoin's governance process best incorporates participation by ful lling the goal of involving all the a ected stakeholders. All the di erent stakeholders can actively participate in the governance process of Bitcoin, compared to the other two permissionless blockchain systems. Anyone is able to submit a proposal, not limiting it to only a speci c group of stakeholders. However, to create fair participation from the stakeholders, Tezos' governance process of limiting participation through a maximum number of votes and available rolls will be utilized. The combination of Bitcoin's and Tezos' governance processes facilitate fair participation and power balance. Furthermore, the type of communication channels that are used in the Tezos Amendment process (such as blogs and messaging) will be incorporated. The motive behind the communication selection is to prevent information fragmentation and is further elaborated in the transparency characteristic. { Rule of law. The utilization of on-chain governance requires the rules and frameworks to be formally de ned. Adoption of Tezos' on-chain governance process, rules and framework pursues this requirement. However, adoption of the amendment process also means adopting its informally de ned hard fork process. As in Bitcoin's BIP, formally de ning hard fork processes can further strengthen the rule of law in the derived fair governance process. Hence, the vagueness surrounding the process is mitigated by formally dened governance rules and frameworks. { E ectiveness and e ciency. E ective governance best approximates people's needs by avoiding hard, soft and user forks in the process. Applying Tezos' on-chain governance that is employed in their amendment process follows the same intent of diminishing those forks. The Tezos Amendment process further in uences the e cient use of resources by preventing other co-existing Tezos blockchains. { Transparency. Integrating Tezos' Amendment governance process results in information that is presented formally, organized, and elaborated on technical as well as non-technical topics through several communication channels. Communication channels which can be perceived by anyone in understandable forms and media (e.g. Medium, GitHub, Twitter, and Telegram messaging) will be incorporated in the new derived fair governance process. { Responsiveness. Tezos' Amendment governance process is the most responsive due to the integration of timeframes throughout the whole governance process (see section 3.3). Embracing similar timeframes in the derived fair governance process will also positively impact participation. { Consensus-oriented. In order to get a wide community consensus, it is essential that all stakeholders are able to express their vote without a bias consent. Incorporating both Bitcoin's and Tezos' governance processes can result in a fair consensus while mitigating bias. Adopting Bitcoin's process to allow any stakeholder to vote creates a wide community consensus. Applying Bitcoin's process of allowing miners as well as users to vote for adopting a proposal supports this wide consensus. Management of proposals is done by a single person (editor) in Bitcoin's governance process, which can lead to bias. By adopting Tezos' on-chain governance to manage proposals, will mitigate this bias. Furthermore, incorporating Tezos' voting limitation of 20 votes per stakeholder avoids cultivating power centralization and bias consent. { Accountability. The stakeholders' accountability in the Bitcoin, Ethereum, and Tezos governance processes can be perceived by their activities represented in the BPMN models. The main di erence in the stakeholders' accountability in the three blockchain systems lies in the fact that only selected stakeholders can vote on the proposal. In the derived fair governance process, both miners and users can vote on-chain for a proposal. This puts the accountability with the voter.

Conclusion

In this paper, wew have presented a proposal for a derived decentralized fair governance model, based on analysis of the strengths and weaknesses of three distinct existing permissionless-based blockchain systems. Decision-making processes in the analyzed blockchain systems (Bitcoin, Ethereum and Tezos) are determined by either o -chain or on-chain governance. The analysis showed that on-chain governance (formally de ned through its code) can reduce de ciencies that were observed in o -chain, and informally speci ed, governance blockchain systems. This reduction is especially apparent in the formally de ned decisionmaking process, a key characteristic of on-chain governance, that positively affects the identi ed fair governance characteristics. The Tezos Amendment process, which utilizes on-chain governance, served as the foundation from which proposal for a more fair process is derived. However, the Tezos Amendment process falls short on three identi ed fair governance characteristics when compared to Bitcoin's BIP process. Thereby, a combination is made of Tezos' Amendment process and Bitcoin's BIP process regarding the fair governance characteristics of participation, rule of law, and consensus-oriented. Participation and consensusoriented characteristics better meet their goal by adding users to participate in the derived fair governance process. Improvement of participation in the governance process can lead to wider community consensus surrounding proposals. Rule of law is further strengthened in the derived fair governance model by formally de ning fork processes. The improved rule of law also has a positive impact on fair governance characteristics such as participation and responsiveness. Future research can build upon the analysis of fair governance characteristics in order to establish a more re ned fair governance model. The derived decentralized fair governance model can act as a steppingstone to develop and adjust the governance process for permissionless-based blockchain systems.

References

1. Moore , J.F. : The Death of Competition: Leadership and Strategy in the Age of Business Ecosystems . Harper ( 1996 )

2. Baldwin , C. , Woodard , C.J.: The architecture of platforms: A uni ed view . Platforms, Markets and Innovation (09 2008 )

3. Mazzucato , M.: The Value of Everything: Making and Taking in the Global Economy . Penguin Books ( 2018 )

4. Nakamoto , S. : Bitcoin: A peer-to-peer electronic cash system . http://www. bitcoin.org/bitcoin.pdf, visited 2021 - 01 - 14 ( 2009 )

5. Kaya , F. , Gordijn , J. , Wieringa , R. , Makkes , M. : Exploring governance in a decentralized energy trading eco-system . In: Proceedings of the33RD BLED econference: Enabling Technology for a Sustainable Society . ( 2020 )

6. Parvizimosaed , A. , Shari , S. , Amyot , D. , Logrippo , L. , Mylopoulos , J.: Subcontracting, assignment, and substitution for legal contracts in symboleo . In Dobbie, G., Frank , U. , Kappel , G. , Liddle , S.W. , Mayr , H.C., eds.: Conceptual Modeling - 39th International Conference, ER 2020 , Vienna, Austria, November 3- 6 , 2020 , Proceedings. Volume 12400 of Lecture Notes in Computer Science., Springer ( 2020 ) 271 { 285

7. OMG: Business process model and notation, version 2.0 . https://www.omg.org/ spec/BPMN/2.0, visited 2021- 01 - 14 ( 2011 ) Object Management Group (OMG).

8. Graham , J. , Plumptre , T. , Amos , B. , on Governance, I. , Canada, P. : Principles for Good Governance in the 21st Century . Policy brief . Institute on Governance ( 2003 )

9. Sheng , Y. : What is good governance (2009) United Nations Economic and Social Commission for Asia and the Paci c .

10. n.a.: BIP 1: BIP purpose and guidelines . https://github.com/bitcoin/bips/ blob/master/bip-0001.mediawiki, visited 2021 - 01 - 14 ( 2011 )

11. n.a.: Governance on Ethereum. https://docs.ethhub.io/ethereum-basics/ governance/, visited 2021- 01 - 14 ( 2020 )

12. Allombert , V. , Bourgoin , M. , Tesson , J.: Introduction to the tezos blockchain . CoRR abs/ 1909 .08458 ( 2019 )

13. Goodman , L. : Tezos | a self-amending crypto-ledger - white paper . https:// tezos.com/static/white_paper- 2dc8c02267a8fb86bd67a108199441bf .pdf, visited 2021 - 01 - 14 ( 2014 )

14. Arluck , J.: Amending Tezos . https://medium.com/tezos/ amending-tezos-b77949d97e1e, visited 2021 - 01 - 14 ( 2018 )

15. Jairam , S.: BPMN models for the governance in Bitcoin, Ethereum and Tezos . https://www.thevalueengineers.nl/ governance-non - permissioned-blockchain/, visited 2021- 01 - 14 ( 2021 )