Blockchains are not companies

The flaws of shareholder voting and blockchain governance

This post mainly addresses blockchain governance. It was written mainly with blockchains in mind rather than blockchain based protocols. However, points made can be applied to both.

On-chain governance discussions often seem to be based on corporate board decision-making. This is a paradoxical issue, given the values behind decentralization. If shareholder voting works for companies then token holder voting must somehow work for all blockchains. This analogy is also used to support the view that token holder voting is not plutocratic. Making this connection allows supporters to suggest token holder voting is both a way of the stakeholders making informed decisions and advancing blockchain norms. In this post I will not only address why I believe blockchains are not comparable to companies in their governance models, but I will also attempt to shed a light on some of the flaws of shareholder voting.

“What is on-chain cryptocurrency governance? Is it plutocratic?” asked Richard Red. He argues that we can not call on-chain voting for blockchain governance plutocratic, because according to his definition:

Plutocracy refers to governance of a society. Blockchains are not societies, but it’s hard to define what the appropriate frame of reference is because blockchains are a new class of socio-technological entity. Relevant frames which spring to mind are States, Companies and Utilities.

Refusing to call blockchains societies underestimates their potential for widespread influence. Red oversimplifies blockchains’ future and reduces them to small networks with minimal dimensions to govern. Blockchains will affect even those not interacting directly with them making social issues an important factor in blockchain governance. For example, social media has profoundly changed our work, even for people who have never opened a Facebook account.

Richard’s post does highlight that it is not easy to define what is being governed on blockchains. There are similarities that can be drawn to states, companies and utilities. This to me is a fair assessment, blockchains are unlike anything we have seen before. Governing them will require the incorporation of various frames in order to build a system that works for this use case. Most of us can agree that blockchain governance is not easy. Therefore, it has always seemed to me that token holder governance is a system proposed by those not interested in doing the hard work needed to solve complicated blockchain governance issues.

Proponents of token holder voting will claim it to be a good solution as it solves for sybil attacks. It provides an easy way to ensure no one has voted twice without having to try and solve for identity. As Vitalik Buterin said in his post “Notes on Blockchain Governance“:

[token holder voting] is an imperfect and unrepresentative signal, but it is a Sybil-resistant one — if you see 10 million ETH voting for a given proposal, you cannot dismiss that by simply saying “oh, that’s just hired Russian trolls with fake social media accounts”.

It is also argued that through a token holder voting system, incentives align. Those with the most tokens inform themselves the most because they have the most to lose.

One of the main differences between blockchains and companies are divergent sets of goals and incentives — and this undermines calls for token holder voting that would mirror a corporate system. These different requirements must be defined.

Disregarding the more classical arguments against token holder voting, various other flaws come to light beyond vote buying that need acknowledgment. The issue is monolithic, towering over a complex system of actors with a relatively primitive governance model.

An interesting article on some token holder voting concerns was written by Philip Daian, titled “On-Chain Vote Buying and the Rise of Dark DAOs”. Some of the concerns in this post become more apparent when we realise that shareholder bribery is an issue that has happened on multiple occasions and is often legal. See:

But shareholder voting works…

Applying capital-oriented governance models (employed by corporations that exist to gather/manage investment and offer ROIs to shareholders) to the blockchain is an inherently flawed idea; the two have vastly differing goals.

This governance model reveals that it is not very efficient, or as successful as proponents claim. Zachariadis, Groen-Xu & Cvijanovic, find that the average turnout for investors lies at 77%. This relatively large voter turnout is mainly due to the fact that most shares are held by institutions, rather than individuals¹. Apple is a strong example of this, where 61.01% of all shares are held by institutions². Is this the kind of governance we want for blockchains, one controlled by organizations whose voting power is derived from wealth or do we need to figure something out which better represents the community forming around the systems we are building? These are questions we need to start answering, and fast — given the increasing number of financial institutions moving towards blockchain investments.

A focus on explicit shareholder voting ignores other implicit ways that influence is put on the governance process. A shareholder sells when they believe their investment is more likely to achieve higher return elsewhere, and they buy when company policy is favourable to their interests. The value of shares can signal satisfaction with the underlying governance. This gauge is probably more useful than most shareholder meetings with a superficial quorum of votes.

It is also important to note that the comparison between shareholder and token holder voting is often poor, as shareholders do not get to vote on important proposals affecting day-to-day business whereas this seems like a goal of multiple on-chain governance models (i.e token holders getting the right to vote on mission critical upgrades).

Not only does it not work because the incentives are different, but there is also a blind trust embedded in shareholder voting. It is assumed that companies govern themselves well and that shareholder voting is an efficient model of governance.

What we must also take into account, is that the process behind shareholder voting is not transparent. First, the management of companies hold quasi-dictatorial power over the shareholder proposals which become subject to vote. Not all proposals have the same chance of being adopted. Moreover, there is significant research suggesting that votes themselves are manipulated. Bach and Metzger discovered in 2016 that there have been 75% more shareholder proposals rejected by a margin of one percent of shares outstanding than proposals that were approved by a similarly narrow margin. They have estimated that roughly 11% of closely-contested proposals that were eventually rejected by voters would have passed if management had not been able to systematically affect the voting results. This shows that management holds vast and exclusive powers over the voting process at big corporations, causing valuable improvements to corporate governance being blocked.

This would beg the question, does shareholder voting work because shareholders do not and can not vote?³ And if this is the case, can we rely on such a model when determining the governance system applied to a blockchain or do we need to dig deeper?