Blockchain Governance: Surveying potential contributions from psychology

What exactly is blockchain governance?

While the word ‘governance’ has a long and rich tradition, it has a unique meaning in the context of public blockchains (see also: open-source governance; open-politics). Derived from the ancient Greek word for steering (kubernan) ships/carts, governance has, in the English language, come to refer to ‘the art of governing’. Conceptually, scholars distinguish governance as a thing (government) from process (governance).

Several sub-strands of governance can help shed light on what it is. Corporate governance attempts both to describe human behavior in complex organizations and to offer practices that improve organizational steering (i.e., coordinating, managing differences and transaction costs). Wider in scope, global governance provides a fairly comprehensive definition:

Governance is the sum of the many ways individuals and institutions, public and private, manage their common affairs. It is a continuing process through which conflicting or diverse interests may be accommodated and co-operative action may be taken (Carlsson et al 1995, p 2)

A useful emphasis here is on solving problems of cooperation. The most recent evolution, ‘modern governance’ (aka. new public management), comes at the heels of a diminishing role of the state resulting from the rise of globalization and market-led privatization (outsourcing) of goods and services. The distinctive feature of modern governance is an approach to regulating society via self-organizing networks (heterarchical), as opposed to by authorities (hierarchical) or economic mechanisms (market) (see Hufty, 2011 for details).

To recap, the historical development of governance moved from bureaucratic government (thing), highly centralized and hierarchical, towards innovative network governance (process-oriented) decentralized and egalitarian. Yet, more context is needed to understand blockchain governance. Geared more towards self-organizing networks, blockchain governance has characteristics of technocracy, highlighting the primacy of software, of engineering and of software engineers (see this report for elaboration).

Thus, from the literature, the following ideas provide a context for understanding blockchain governance. Governance includes:

• human behavior in complex organizations
• solving problems of cooperation and coordination
• conflicting and diverse interests
• self-organizing networks
• decentralized and egalitarian
• primacy of software, engineering and software engineering

What are practitioners saying?

Cryptoeconomics is the defining feature of what makes public, permissionless blockchains interesting as Josh Starks rightly points out. It’s the use of cryptography and economic incentives to produce consensus among distributed network nodes that makes public blockchains special.

Nevertheless, cryptoeconomics (machine consensus) is only part of the picture. On a continual basis, the software that the network is running on needs to be updated, improved and fixed; this requires social consensus or governance.

In a recent presentation, Vlad Zamfir laid out a description of blockchain governance as involving three inter-related strands of activities:

1. Governance of decision-making processes
2. Coordinating stakeholder participation around those decision-making processes
3. Managing the legitimacy of decisions

At present, all three cluster of activities require human involvement.

Zamfir goes into details on seven participants (and institutions) who are directly involved in either upgrading, improving, fixing bugs/vulnerabilities or, potentially making irregular state transitions on Ethereum. This includes: Full nodes, Miners, Geth/Parity implementations, the ‘All Devs Call’, Ethereum Improvement Proposal (EIP) and the Roadmap.

Vlad Zamfir @ Deconomy (April, 2018)

Note: If we were to include indirect participants/institutions, then the Ethereum subreddit, Ethereum Gitter chat room, and institutions like Ethereum Foundation and Ethereum Community Fund could arguably be included.

Vitalik Buterin, in describing the coordination model of governance, distinguishes between two layers. First, there’s the deciding layer (layer one) where nodes, validators, miners or any user make decisions about what software to run; and these decisions must be coordinated with other users (e.g., you don’t want to run a version or implementation that no one else is on).

On one hand, people listen for information about what their peers are going to do (as well as their own incentives, guesses, beliefs, expectations and tacit knowledge) and act on that information. On the other hand, people are reading/reacting to signals from coordination institutions (layer two) and this also factors into the decision-making process to the extent which the signals are viewed as legitimate (see this post for details).

Given that all participants are physically distributed, each having to figure out which software their peers are running, while also interpreting signals from various coordination institutions, blockchain governance represents a coordination challenge.

For example, there is a fair amount of social coordination involved in even proposing one EIP, especially for changes to the core protocol (ref EIP workflow). The author of an EIP is motivated in getting her EIP accepted. She must send out signals to the community to gauge alignment (informal vetting). The editor is motivated to screen that EIP, among many others, for quality, suitability and compatibility with the Ethereum protocol; the editor is authorized to exercise human judgment.

In his presentation of Blockchain Governance 101, Zamfir, in discussing the legitimacy of Ethereum Governance, distinguished between decisions around three sub-processes including:

1. Maintenance and recovery
2. Technical upgrades (EIPs)
3. Ethically-motivated hard forks

While all three require coordination, Zamfir maintains the third process as particularly challenging.

Cooperation, Competition and Conflict

Social psychology has a rich tradition of studying conflict situation (Deutsch, 2011). Of particular interest to researchers was that conflict situations involved mixed motives with both cooperative and competitive forces intertwined (Schelling, 1958).

In drawing on game theory, researchers have argued that knowledge of coordination games could readily be applied to understanding organizational life. Furthermore, coordination games are often of mixed motives (Camerer & Knez, 1997).

Thus, putting on a social psychology lens, coordination games inherent in blockchain governance could be framed as situations where parties have mixed-motives with conflicting and diverse interests (see global governance definition above).

And yet, not all conflict situations are equal.

There are varying degrees of conflict, with the majority of cases easily resolved. For example, according to Zamfir, the complaints around maintenance, recovery and technical upgrades (EIP work flow) have been minimal.

Nevertheless, what makes blockchain governance so important are the situations where conflicts involve ideology, emotions and core beliefs, which make them controversial. These conflicts are what Zamfir refers to as ethically-motivated hard forks (e.g., DAO hard fork; Parity hack). These situations conjure up questions around procedural fairness and power. More crucially, these conflict situations could potentially escalate and split the community into factions, or lead to an impasse, or ultimately a hard fork.

What matters for blockchain governance are processes to minimize escalation of conflict; and, moreover, finding cooperative ways to resolve such conflict when they occur. Here, insights gained from decades of social psychology research may have something to contribute.

The basic insight from this body of knowledge is that in social situations, the structure of people’s incentives (goals), can be characterized as cooperative or competitive (most often a mix of these two types)— with implications for the following:

• Effectiveness (or ineffectiveness) of communication
• Friendliness and helpfulness (or obstructiveness) of relations
• Coordination (or lack thereof) of effort
• Feelings of respect (or disrespect, critical rejection)
• Defining conflict as a mutual problem to be solved collaboratively (or competitively with coercive tactics, in a power struggle)

When conflicts inevitably arise, constructive processes of conflict resolution are similar to cooperative processes of problem solving, while destructive processes of conflict resolution, are similar to competitive processes (Deutsch, 2011).

There are several implications for blockchain governance. First is the importance of developing a cooperative (win-win) orientation. While obvious at first glance, this is more than simply telling people to adopt this orientation, but it involves examining incentive structures in the system so they do not foster win-lose relationships. Moreover, once conflict arises (e.g., break down of consensus), it’s important to reframe the situation as a mutual problem to be resolved through cooperative efforts. Perhaps most importantly, the question becomes:

Can we develop governance norms that incorporate norms of cooperation?

Trust

Research has established that an important correlate of cooperation is trust (Balliet & Van Lange, 2013). In social dilemmas — situations of conflict between self- and collective-interest — the positive correlation between trust and cooperation is stronger when there is a larger degree of conflict. For example, in Prisoner’s Dilemma, there are four outcomes: Temptation(T), Reward(R), Punishment(P) and Sucker(S) (Rapoport, 1967). Rapoport’s index of cooperation is:

R-P / T-S

The larger the distance between Temptation (T) and Sucker (S), the smaller the index of cooperation. A lower index of cooperation means the temptations to behave opportunistically (competitively), with no regard for mutual benefit/cooperation, is higher.

Numerous studies have found that in situations of higher conflict (lower index of cooperation), there is a stronger positive association between trust and cooperation (Balliet & Van Lange, 2013). For blockchain governance, this implies that when we have potentially high conflict situations, like controversial ethically-motivated hard forks, if we want people to cooperatively and constructively resolve conflict, we want to foster norms of trust.

Thus, the irony of all this is that trust, cultivated over time, off-chain, may be required to facilitate trustlessness (driven by cryptoeconomics) on-chain.

Justice and Legitimacy

A critical part of blockchain governance is managing the legitimacy of the decision-making processes. On one level, legitimacy ensures that signals from coordinating institutions are effectual, as Buterin suggests:

You can think of coordination institutions as putting up green or red flags in the air saying “go” or “stop”, with an established culture that everyone watches these flags and (usually) does what they say — Vitalik Buterin.

In addition, Zamfir points out that legitimacy of the decision-making process promotes (user/voter) engagement, while preventing (user/voter) apathy and undesirable outcomes like hard forking. What gives a process its legitimacy is procedural fairness.

Conflict scholars have studied the effects of injustice on conflict. The insights from this body of knowledge could potentially elaborate our understanding of the concepts like legitimacy and procedural fairness in blockchain governance. The types of injustices that have been studied include (Deutsch, Coleman & Marcus, 2011):

• Distributive injustice: criteria leading to perceived unfair outcomes
• Procedural injustice: fair treatment in making and implementing decisions that lead to the outcome
• The sense of injustice: factors that determine whether an injustice is experienced as such
• Retributive and reparative injustice: how to repair the community once justice has been violated
• Moral exclusion: who is thought to be entitled to fair outcomes
• Cultural imperialism: when a dominant subgroup imposes values, norms and customs on another group

The conflict literature has found that long-term conflict situations that develop into oppression are characterized by various forms of injustices. Injustice is thus antithetical to fostering a cooperative orientation for collaborative problem solving. In conflict situations, processes or outcomes which are perceived as unjust would likely lead to an unstable resolution. Extending to blockchain governance, a lack of procedural fairness would not only make people unsettled/upset, it would be akin to sowing the seeds to further conflict.

Some implications of this research are:

1. Humanization of the other: Conflict and injustice is often accompanied by dehumanization. In distributed communities where different stakeholders may get little opportunity to interact on a face-to-face level, this may be particularly risky (note: look no further than the flame wars on twitter to see how this can escalate or the ongoing animosity between bitcoin core and bitcoin cash). In the face of potential conflict, it’s important to curb propensities to dehumanize.
2. Fair rules for managing conflict: It may be helpful to anticipate conflict, particularly if we view conflict management as inherent in governance processes, creating rules/processes for managing disagreement or breakdown in consensus beforehand will be helpful.

A blockchain governance that seeks to develop cooperative norms, constructive conflict resolution and legitimacy of decision making will prioritize justice.

Decision-Making & Consensus

Organizational scholars have studied decision-making in groups extensively (Lunenburg, 2011). In consultative decision-making, a clear leader consults with group members, while in democratic decision-making, the group, sans a clear leader, is empowered to make decisions. Blockchain governance, will naturally lean toward the latter, as public blockchains seek to avoid reliance on a central authority / leader figure as typical in most organizations.

In a traditional organization context, consensus decision-making, lies between consultative and democratic. Not requiring unanimity, consensus happens when participants believe their point of views and interest are represented fairly and thoroughly and the resulting decision — even if they don’t agree — is one they believe, overall, is in the best interest of the organization (Hartnett, 2011).

Building on the discussion above, consensus decision-making is consistent with building cooperative relations, which would help foster constructive resolution, in the event of conflict.

However, there are well documented obstacles to group decision-making (negatively affecting consensus) including: groupthink (Janis, 1982), group polarization (Bordley, 1983), escalation of commitment (Whyte, 1993) and the abilene paradox (Harvey, 1974). Moreover, other barriers to group consensus decision-making include: group size and group diversity. In fact, some researchers have explicitly recommended making groups homogenous (limiting diversity) and cultivating exclusivity to increase group cohesion, and by extension, quality of decision-making (Luthans, 2011). Unfortunately, this appears to be antithetical with the philosophy of the Ethereum and other blockchain projects.

Nevertheless, several techniques have been developed to help groups make quality decisions including: brainstorming, nominal group techniques, delphi technique, devil’s advocacy and dialectical inquiry (Lunenburg, 2011).

In the context of blockchain governance, insights around group consensus and decision-making could be applied to structure discussion prior to, or in anticipation of, potentially controversial topics (i.e., ethically motivated hard forks). Effective application of these techniques could also help cooperative problem-solving even in cases where consensus is not reached.

Organizational Culture & Climate

What makes blockchain governance interesting, yet tricky, is what Zamfir describes as the common knowledge which serves as shared norms that allow people to act as if those norms were fact and could be taken for granted. This allows norms to be enforced and facilitates complex coordination of multiple stakeholders (i.e., when participants receive signals from coordinating institutions, common knowledge allows people to coordinate without the need to re-learn new rules from scratch). What’s tricky is that common knowledge and external governance structure is really difficult to change.

The way common knowledge is described makes it very similar to the construct of organizational culture which is:

A pattern of shared basic assumptions learned by a group as it solved its problems of external adaptation and internal integration — Schein (1992)

Colloquially, organizational culture is ‘the way we do things around here’.

A similar construct is organizational climate, which is the shared perception and experiences — a gestalt — of policies, practices, procedures and what is rewarded, supported and expected in organizations (Schneider et al., 2017).

Organizational psychologists have extensively researched ways of classifying, comparing, analyzing, influencing and changing both culture and climate. In the context of blockchain governance, if there was desire to actively influence common knowledge, this body of knowledge could provide insight.

However, researchers have identified the primary role that leadership plays in influencing both organizational culture and climate, implying that an active approach to ‘managing’ blockchain governance may require some degree of political centralization.

Conclusion: Some Open Questions

Is most of blockchain governance community consensus forming? What are the modes of decision-making?

Community consensus decision-making (leaning towards democratic decision-making) may take more effort and not be as efficient as other modes, but it helps foster legitimacy and cooperative norms. A distinction is made between the mode of consensus decision-making and the process of arriving there. As a community, if we can build tools and mechanisms that allow people to discuss their views and feel like their views are represented fairly, then sufficient goodwill should be developed. In cases where consensus is not reached, the governance norms that are built around cooperative/collaborative problem-solving should help prevent perceptions of illegitimacy and revolution.

What should devs do when there’s no clear consensus about a core EIP or any core protocol change?

Viewing governance as conflict management, mechanisms and processes could be built prior to, in anticipation of a lack of clear consensus around core EIPs/protocol changes, to ensure that conflicts are resolved constructively. These mechanisms would foster norms around cooperation, trust and justice.

The literature around collective intelligence, that have studied peer production systems, may also help shed light on exactly how these mechanisms could be built.

Should blockchain governance be active?

Organizations spends a lot of resources to get people to identify with the organization’s goals. I’d argue this organizational identification already exists for Ethereum (and other blockchain projects), but it would be interesting to see this is something that, moving forward, is actively and explicitly cultivated.

Perhaps there are levels of activity. Actively managing common knowledge (i.e., climate/culture) may require input from leadership. Thus active blockchain governance is, to a degree, associated with political centralization. As long as Ethereum goes with off-chain governance, some level of active governance may be desirable.

The attention paid to fostering norms of cooperation, trust and justice may go a long way towards offsetting any potential downside to a degree of political centralization. During times of crisis people will reflexively look towards leadership. At this point, we do not know if on-chain governance will be negatively perceived as a leadership void in the event of a crisis.

Should there be a legitimate process for doing irregular state transitions?

A process for discussing, coming to consensus, handling lack-of-consensus and then finally deciding once consensus is reached should help any process for handling irregular state transitions achieve legitimacy, while building cooperative norms and trust.

How should we handle conflict?

Hopefully, I’ve made a case that viewing blockchain governance as an ongoing process of conflict management is productive. While the benefits may not be readily apparent for technical upgrades and maintenance, the norms associated with cooperative and constructive conflict resolution should be helpful in handling ethically-motivated contentious (potential) hard forks.

Limits of Psychology

There are important caveats to the generalizability of social and organizational psychology insights for blockchain governance. Although social psychological insights extend to situation where people are influenced by the ‘actual, imagined or implied’ presence of others, only recently have researchers examined contexts were people are interacting, for the most part, virtually (i.e., social media, messaging apps, chat rooms, bi-monthly video chats, occasional in-person contact at conferences).

Thus, it is an open question, the extent and magnitude that these insights apply to networks of distributed individuals whose relationship does not resemble any type of social and organizational membership we’ve seen prior to 2009.

Finally, while the bulk of this article focuses on potential contributions from social and organizational psychology, in reality, the challenges of blockchain governance is multi-disciplinary. Other areas for fruitful exploration include the emerging, interdisciplinary body of knowledge around collective intelligence and behavioral economics.

References

Balliet, D., & Van Lange, P. A. (2013). Trust, conflict, and cooperation: a meta-analysis. Psychological Bulletin, 139(5), 1090.

Bordley, R. F. (1983). A Bayesian model of group polarization. Organizational Behavior and Human Performance, 32, 262–274.

Camerer, C., & Knez, M. (1997). Coordination in organizations: A game-theoretic perspective. Organizational decision making, 158–188.

Deutsch, M. (2011). Cooperation and competition. In Conflict, interdependence, and justice (pp. 23–40). Springer, New York, NY.

Deutsch, M., Coleman, P. T., & Marcus, E. C. (Eds.). (2011). The handbook of conflict resolution: Theory and practice. John Wiley & Sons.

Hartnett, T. (2011). Consensus-oriented decision making: The CODM model for facilitating groups to widespread agreement. Philadelphia, PA: New Society Publishers.

Harvey, Jerry (Summer 1988). “The Abilene Paradox: the Management of Agreement”. Organizational Management. American Management Association. 17 (1): 19–20.

Janis, I. L. (1982). Groupthink: Psychological studies of policy decisions and fiascoes (2nd ed.). Boston, MA: Houghton Mifflin.

Lunenburg, F. C. (2011). Decision making in organizations. International journal of management, business, and administration, 15(1), 1–9.

Rapoport, A. (1967). A note on the” index of cooperation” for prisoner’s dilemma. Journal of Conflict Resolution, 11(1), 100–103.

Schelling, T. C. (1958). The strategy of conflict. Prospectus for a reorientation of game theory. Journal of Conflict Resolution, 2(3), 203–264.

Schein, Edgar (1992). Organizational Culture and Leadership: A Dynamic View. San Francisco, CA: Jossey-Bass. p. 9

Schneider, B., González-Romá, V., Ostroff, C., & West, M. A. (2017). Organizational climate and culture: Reflections on the history of the constructs in the Journal of Applied Psychology. Journal of Applied Psychology, 102(3), 468.

Whyte, G. (1993). Escalating commitment in individual and group decision making: A prospect theory approach. Organizational Behavior and Human Decision Processes, 54, 430–455