– May 16, 2017
Share:

Bitcoin’s system of radically decentralized governance is facing perhaps the biggest test in the digital currency’s history. As my colleague Patrick Coate described, the code used to run Bitcoin’s blockchain database needs to be updated to transmit more information faster. However, there are two major proposals on the table for how to change the code. The controversy has raged for months and may even result in Bitcoin splitting into two currencies.

Many of Bitcoin’s greatest strengths come from its decentralized nature, but if there were somebody at the center of its network making decisions, the current problem would look very different. The governing body would choose one path, and those who didn’t like it could sell their Bitcoins. This structure would resolve the uncertainty much faster, and not result in a split that would likely scare off some potential Bitcoin users.

The decentralized structure of Bitcoin is not, alas, a free lunch. Decentralization is a tradeoff with pros and cons relative to a more traditional structure. This article will consider those pros and cons. But first, let’s look at how decisions are currently made in the Bitcoin network, and what a more centralized structure might look like.

Bitcoin vs. Bitcoin Inc.

Decisions like modifying Bitcoin’s code currently involve interactions between developers (anyone who wants to write open-source code and propose to the Bitcoin community that it be adopted) and miners (those who run the code behind the blockchain database and are rewarded with new Bitcoins). In practice, the blockchain database must “fork,” meaning it branches into two blockchains, one with the change and one without. Each miner has control over the code they use. They could in theory make a unilateral change, but that would stop their copy of the database from being linked to all the other “nodes.” For a change like the one currently on the table to be adopted, 95 percent of mining nodes must approve and run the new code. (Miners and developers can communicate in online forums.)

But suppose these decisions were instead up to a centralized entity that we’ll call Bitcoin Inc. Bitcoin Inc. is a hypothetical private for-profit firm that has some ownership and control over the network. Assume all code is still open-source and can be written by third party developers, and that the system is still run by open mining (i.e., this is still a public blockchain rather than a private, permissioned one). However, decisions about changes to the code, rules, and dispute settlement lie with Bitcoin Inc. The company could get its revenue by either doing some mining itself or charging a miniscule transaction fee. These alternatives are interesting to consider but don’t matter for the exercise.  While this system involves more centralized control, it is still private, lacking any government management of the currency.

In what follows, I’ll make the case first for Bitcoin Inc., and then for the currently decentralized Bitcoin. Bitcoin’s current governance structure prevents undue influence from any single individual or organization, but does so at the expense of not being able to quickly innovate or respond to unforeseen events.

The Case for Bitcoin Inc.

It is impossible for Bitcoin stakeholders to fully plan for every future event. Any coordinated action that these stakeholders currently take is governed by the implicit contracts between them that emerge from the code and the economic incentives it creates. As economist Ronald Coase famously observed, one of the reasons firms exist at all is that in a world of uncertainty, writing enforceable contracts between parties for every possible future state of the world is impossible, and even if it were possible, prohibitively time-consuming. Firms instead put some discretionary power and control in the hands of human beings, who can respond to new information as it arrives and have incentives sufficiently aligned with the firm. In a world with Bitcoin Inc., the firm would make a call on how to change the code, and individuals could decide whether to use the currency, like every product in the market. This system would also be more efficient in deciding whether to reverse allegedly fraudulent transactions. This issue resulted in a split in Ethereum, another cryptocurrency. The inability of these radically decentralized systems to decide whether to reverse a fraudulent transaction is evidence for Coase’s observation and suggests just how difficult it is to fully plan for even predictable outcomes.

This difficulty with rapid coordinated action leads to an undeniable inertia in currencies such as Bitcoin. When Bitcoin was first conceived in Satoshi Nakamoto’s 2008 paper, it was at the technological cutting edge. Nine years later, many argue that Bitcoin has been eclipsed by other cryptocurrencies as state of the art. Bitcoin cannot readily adopt the open-source innovations of other cryptocurrencies. To do so requires the cumbersome “forking” process described above. This process seems unnecessarily risky and time-consuming, and renders even small changes in the code a big deal. On the other hand, a firm like Bitcoin Inc. would have an R&D department, freely experimenting with new programming innovations evolving from both internal work and other digital currencies. Should management conclude that a change is due, it can simply make the change, and miners or users, just like customers of a restaurant that changes its menu, are free to go elsewhere if they don’t like the change.

The event of a fork turning into a full split devalues both new currencies. Currency adoption and use is fueled by network effects: the currency is more valuable when more people use it. In a hypothetical 50/50 split, each new currency would have only half the users it did before. In addition, we are still at a point where most people do not understand or fully trust cryptocurrencies. Such apparent complexity might delay their adoption.

Bitcoin’s radically decentralized structure has a clear weakness in being unable to respond to new information in a rapid and coordinated manner. Implementing change is a very complicated process, but is it a worthwhile tradeoff to reap the benefits of decentralization?

The Case for Decentralized Governance

Bitcoin advocates talk about the decentralized system having no single “point of failure.” They usually mean that because the blockchain database is distributed rather than having any master copy, a hacker would have to attack an almost prohibitively large number of network nodes to change the data. This important feature is maintained in the world of Bitcoin Inc., since the database would still be fully distributed. But the point-of-failure concept is also highly relevant for control of a currency and the ability of that control to be coopted. Currently, no one individual or organization has the power to unilaterally change the code or rule on the validity of a transaction. The fact that no individual or organization can unilaterally change Bitcoin’s code or rule on the validity of a transaction protects it from government interference, corporate crime, or agency problems stemming from differing individual incentives. More centralized control by Bitcoin Inc., just like any large corporation in a similar situation, would open the system to all these issues.

There are also advantages in Bitcoin’s decentralized, bottom-up process for proposing and implementing changes. While slow and unwieldy, the process may allow creative and effective solutions to evolve that Bitcoin Inc. would never think of from scratch. In some ways this is like the free market generating innovations that would elude any central planner. It also illustrates why splits like the one that happened in Ethereum and may happen to Bitcoin aren’t necessarily bad. Users are free to choose between the version of the currency with the change and the one without, rather than having a central entity guess which choice is best.

Finally, lack of change itself may be a desirable feature for a cryptocurrency. One of the biggest issues with the fiat currencies widely used today is that their centralized controlling entities, national governments, can manipulate the currency anytime they want. In a system like Bitcoin where change is incremental and not sudden, users can have far clearer expectations about the currency into the future.

Other Governance Models

Bitcoin’s current structure and the hypothetical Bitcoin Inc. are only two of many possible systems of governance for a cryptocurrency. In the past several years, the market has produced hundreds of so-called altcoins (non-Bitcoin cryptocurrencies), each with its own characteristics and means of governance.

Ethereum, the largest of the altcoins, has the nonprofit Ethereum Foundation at its center. The foundation does not have the power to unilaterally change code or reverse transactions, but does carry a large amount of influence in the Ethereum community. Ethereum’s creator, Vitalik Buterin, and his colleagues on the foundation team constitute a voluntary coordination device for Ethereum’s miners: Buterin can use the influence and trust he engenders as the currency’s creator to rally support for changes to Ethereum’s code that might take longer for a system like Bitcoin’s to implement. When he proposed the “hard fork” to reverse a fraudulent transaction, the intention was for all Ethereum miners to go along with the decision. While 85 percent of Ethereum’s miners agreed, the remaining 15 percent did not support the action, resulting in a full split. Ethereum Classic was born, and what was once a single digital currency became two separate currencies running on separate blockchains. Buterin and his foundation’s influence goes only so far.

Another solution is to formalize, through code, means for stakeholders to vote on or approve changes short of splitting the currency. Suppose Bitcoin’s miners (and potentially also users) could vote to accept or reject a proposed coding change, and if a prespecified percentage of voters approved, the change would automatically go into effect over the entire network. Tezos, developed in 2014, is one example of such a system. Cryptocurrencies could go even further in the direction of a representative democracy, electing members of the community to serve limited terms in which they would have some decision-making power over the code and transactions.

Ready for Prime Time?

Suppose a global crisis of confidence in fiat currencies erupted tomorrow. Would Bitcoin as currently constituted be able to fill the vacuum and become a global currency? The answer is almost certainly no. The current controversy that may lead to a split is about scalability on a much smaller level than would be required to replace the dollar. And scaling up, especially against the backdrop of an economic crisis, would require exactly the rapid change and innovation with which Bitcoin struggles.

But what if the timetable to replace fiat currencies was 50 years rather than overnight? In this scenario, the ultimate winners among cryptocurrencies have probably not yet been invented, and the messy process of trial and error inherent in Bitcoin’s current system may be desirable. If no single cryptocurrency needs to be ready for prime time anytime soon, one or more splits in Bitcoin could yield the type of market-based evolutionary innovation discussed above. Governance of Bitcoin and other cryptocurrencies is still very much a work in progress.