†The Political Economy of Blockchain Governance. [PDF]
Barton Lee (University of New South Wales), Daniel J. Moroz (Harvard University), David C. Parkes (Harvard University)
(† Hot Topics accept, abstract & review summary to journal)
Paper summaries from the reviewers:
“The paper offers a game theoretical model to understand the downstream consequences of policy proposals in decentralized environments. It shows how the three possible forking outcomes (contentious, non-contentious, null fork) can be achieved depending on network effects, policy preferences, and the friction introduced by exchanges. A nice, convincing model, understandable even for those without in-depth expertise in mathematical modeling.”
“This paper presents a political economy model of blockchain governance which can be used to study how forks are implemented. Most interestingly, their results explain some counter intuitive outcomes in blockchain governance, such as the movement from PoW to PoS protocol in Ethereum.”
“This paper presents an elegant model which provides new insights into how forks are implemented. The model analyzes how different types of forks (contentious, non-contentious, null) are proposed by proposers and adopted by users under different conditions (in the presence of market frictions and network effects). Authors use their model to demonstrate the market conditions under which more or less contentious forks arise.”
“This article develops a theory of blockchain governance using tools from formal political theory. The software underlying blockchain projects are frequently updated (’forked’) to implement new policies, and these forks are the subject of our inquiry. The author(s) investigate the way in which the preferences of policy proposers and users influence which policies are actually implemented. They consider users’ utility to be driven both by a preference for a particular policy as well as preference for network-effects. They describe several types of forks and identify the strategic conditions necessary for each. Their results highlight the influence of market frictions created by cryptocurrency exchanges in promoting non-contentious forks and they show that network-effects can incentivize more extreme policy reforms. They then discuss how these results provide an explanation for some counter-intuitive phenomena they observe in live blockchain systems.”
“Related work for context includes: Beck, Roman; Müller-Bloch, Christoph; and King, John Leslie (2018) "Governance in the Blockchain Economy: A Framework and Research Agenda," Journal of the Association for Information Systems: Vol. 19 : Iss. 10, Article 1”
Comments on the strength of the paper:
“This article deals with a very important topic. It is the first to address it, to the best of my knowledge. For that, it uses a clear set of data and is very persuasive. It introduces new distinctions and classifications.”
“What a great article! Perhaps the result (i.e., the fact that "the governance structure of blockchain applications and the cryptocurrency environment more broadly influences the incentive to fork and the types of policy changes that will be proposed") could be exploited a bit more. It has tremendous policy proposals. In any case, this article is a true game-changer.”
Critiques & author responses:
Comment 1:[The work should include...] more data to test the different models.
Response: We agree; however, due to data constraints it seems difficult to run such a test rigorously. For example, a rigorous test would likely require measuring the distribution of users’ preferences on a given currency, which is not feasible. As is standard in applied theory papers in economics and political science, we feel we must settle for less rigorous and more anecdotal evidence as “validity” checks for our theory and models.
Comment 2: Re: proposal 7 (spoiler effect) - discussion assumes that proposer knows the value of their proposal ex ante - what happens in light of incomplete information or mistakes in valuation. Also when you refer to "many proposers," it would be helpful to distinguish between multiple proposers who propose different proposals and multiple proposers who propose the same proposal as the latter can explain the network effects.
Governance is "sticky" (hard to change once implemented) and thus starting point matters as much as (if not more than) midstream changes. Paper should address the question of how the starting point is set and by whom in addition to midstream amendments.
Response: Yes, this is a great point; our model and paper is agnostic about the starting point (status-quo) policy. This is an interesting avenue for future research that we will add to the paper’s discussion / conclusion section. One first guess may be that the status-quo policy is approximated by the median of the users’ preferences (since this is the Condorcet winner). When the status-quo coincides with the median, our model suggests that non-contentious forks should be less likely and only contentious or null forks arise. However, this first guess is perhaps too naive since a currency’s users can change while the policy remains unchanged. For example, when a currency is created, suppose that the median of some initial set of users’ preferences are used to construct the status-quo. As users join and/or leave the currency it is likely that this status-quo policy becomes a worse approximation of the new median. Incorporating such a feature in to our model and, in turn, analyzing how our model’s predictions (possibly) change is an interesting research direction. Thanks!
Comment 3: I see a weak connection to political economy models, despite the claims. I have two reasons to say that. On the one hand, there are many rather different types of discourses, all covered by the term "political economy". Formal, microeconomic utility modelling is one, rather narrow interpretation of the term, which may be confusing for those who expect a more political or sociological approach to the analysis. On the other hand, the connection to the cited political economy literature (Hotelling, and Downs) is mainly the Median Voter Theorem, but there is no substantial explanation how a theory, with its own assumptions, formulated in the context of democratic elections applies (or not) in the context of user choices in a quasi-market scenario. It is also a weaker than ideal connection to real world issues. The effort to connect the theoretical findings to the outcomes of real world governance issues is a very nice addition to the paper, but at this moment it is not done with a sufficient rigor. A more convincing, and systematic link should add greatly to the value of the paper.
Response: We agree, it is unfortunate that “political economy” has different meanings in different context. Nonetheless, we prefer to use this term. It is in line with many papers, for example:
–“The political economy of the Kuznets Curve” Acemoglu and Robin-son, Review of Development Economics, 6(2), 2003
–“A model of the political economy of the United States” Alesina, Londregan and Rosenthal, 87(1), 1993.
–“The political economy of government debt” Alesina and Passalac-qua, Handbook of Macroeconomics, volume 2