Updated: December 29, 2025
The relevance of the theory of the firm for blockchain-enabled decentralization is immediate and promising for researchers
Blockchain-enabled systems introduce two fixes that theoretically replace the firm: the cost of monitoring and the hold-up problem
Two "blind spots" of the blockchain narrative: ownership structure and conflict-resolution
Introduction
Recently sitting in the panel and listening to the conversation between Youngjin Yoo, Jungpil Hahn, Ola Henfridsson, Hanna Halaburda, Daniel Obermeier, and Christoph Mueller-Bloch, I heard an interesting call for action: “I wish someone took the theory of the firm and analyzed its suitability to decentralized systems.”
So, here I am summarizing my read of the theory of the firm and related academic works and reflecting on how decentralized, blockchain-enabled systems fit or extend the current theory.
Theory of the firm recap
The theory of the firm is built on the Coasean idea of the inefficiency of market structures. The core reason for such inefficiency – transaction costs – stems exactly from the decentralization, or disaggregation, of the market network. Searching for decentralized suppliers creates inconvenience (cost) and probability of a not-perfect match (cost); negotiating and writing contracts is prone to power imbalances between parties and, as a result, the possibility of non-fair bargains; enforcing contractual terms also resides on power, and small players may not be eligible to perform this function due to prohibitively high cost; monitoring contract execution and third-party performance imposes similarly prohibitive cost. Therefore, to avoid all these costs and frictions, agents group in hierarchies that we call firms. The core premise of the firm hierarchy is to do internally what would be too costly to contract for externally; outsource what markets can do cheaply.
Williamson further extends the theory by introducing the concept of governance structures and his famous three types of governance: market, hierarchy, and hybrid. Williamson extends the narrative when he considers attributes that determine the hierarchization of economic relationships between agents, such as asset specificity, uncertainty and complexity of transactions, frequency of transactions, and agent opportunism. The core tension lies in the intensity of each of these factors and the resulting cost of contract enforcement. For example, when asset specificity and uncertainty are high, market contracting risks hold-up and leads to costly renegotiation. To avoid this tension, agents choose to vertically integrate, or internalize these relationships under hierarchical employment contracts, which allow stricter enforcement of contractual terms and enable a punishment system for non-obedient actors and exclusion of such agents from future relationships.
Grossman–Hart–Moore shift the focus to a solution for incomplete contracts – residual rights of control – which is based on ownership of assets. This position directly suggests single-agent ownership as the most efficient model for conflict resolution and introduces another tension: the cost of collective ownership and preference aggregation, discussed extensively by Arrow and economists.
Even when firms hierarchize, there is a host of problems inside their hierarchical structures, e.g., the free-riding problem. The core of these tensions is best conceptualized as a principal–agent problem, which implies that owners and hired agents have different incentive structures and information sets that are not fully available to each counterparty. Therefore, the contractual problem remains: to design incentive systems that promote the truest signaling of agents’ qualities and extract the fairest reward for this skillset from the principal.
What’s new about blockchain-enabled systems
Reflecting on the prior research and theories explaining the existence of the firm above, I came up with two "fixes" for the problems that blockchains may substitute the firm, and two "blind spots" in the narratives about blockchains.
Blockchain introduces electronic systems and automation of enforcement, monitoring, and search (Internet).
Smart contracts address the hold-up problem in non-hierarchical (non-firm) environments.
The narrative about decentralized systems does not specifically suggest that ownership should be collective (or private, for that matter).
While principal–agent contracts do not arise in communal systems, it may be worth thinking about conflict-resolution mechanisms for decentralized systems.
Let's elaborate.
Putting the points above together, blockchain-enabled systems (as electronic systems) clearly alleviate some of the classic Coasean frictions, especially around monitoring and tracking, because digital infrastructure can record, verify, and coordinate at scale. This aligns with the first fix: blockchain introduces electronic systems and automation of enforcement, monitoring, and search. However, that does not mean the system becomes frictionless: in practice, these tools are costly to develop and maintain, and they constantly have to handle exceptions (and then operationalize those exceptions into the workflow). So even electronic systems remain “institutional machines” with real maintenance burdens rather than costless pipes.
At the same time, the story of “lowered search costs” is complicated. Yes, the Internet reduced search costs dramatically, and it is itself a decentralized system governed through non-profit and infrastructure arrangements maintained by a bunch of individual firms, producing something close to a public good. Yet, precisely because search becomes cheap, the market becomes crowded, and signals of quality can be easily manipulated. AI chatbots may remove part of this crowdedness, but they also remove choice, and the decision of whether to introduce bias into the averaged-wisdom output belongs to the providers of these common intelligence tools.
This then leads to my broader point: transaction costs don’t disappear; they relocate. Even when monitoring and enforcement become cheaper (or more automated), costs emerge somewhere else, in dimensions we didn’t even think about before. In my previous article, I reflect on the loss of the signaling value in the market driven by AI agents (due to the lower cost of creating content in large volumes and mimicking good-quality content). So the hypothesis goes beyond “blockchain doesn’t eliminate transaction costs” and toward something stronger: the sum of costs seems to have a lower boundary that stays roughly constant, independently of how the contributing elements shift proportions (like the second or first law of thermodynamics). The second fix fits here as well: smart contracts address the hold-up problem in non-hierarchical (non-firm) environments, yet the up-front specification and exception-handling costs can reappear elsewhere in the system.
Finally, there is another aspect I find both most promising and underdiscussed (imho): blockchain-enabled electronic systems operate on the premise of private ownership of resources. Ideologically, we are used to thinking about the communal aspect of decentralization, while forgetting to reflect on the efficiency introduced by the complementarity of agents’ toolsets for the creation of greater value (reflection here to the theory of relative competitive advantage in international trade). At the same time, this is exactly where the first blind spot matters: the narrative about decentralized systems does not specifically suggest that ownership should be collective (or private) for that matter, and this ambiguity shapes how we interpret what “decentralization” is supposed to accomplish. This also implies tangibility of interactions between agents in decentralized systems, i.e., not only casting votes on intangible decisions, but contributing scarce and complementary resources. From this perspective, scarce information innate to individual agents and their creative potential becomes a valuable complementary input. But “agent enforcement with common knowledge” does not produce this complementarity and thus, hypothetically, will not lead to stable, value-enhancing decentralized collaboration.
The second blind spot follows naturally: while principal–agent contracts do not arise in communal systems in the standard way, it remains worth thinking about the conflict-resolution mechanisms for decentralized systems, especially when incentives are diffuse, monitoring is automated but incomplete, and exclusion/punishment is harder to implement without a centralized arbiter.
Conclusion
While electronic, blockchain-enabled systems introduce a bunch of technical tools addressing the challenges of economic exchange that previously were conceptualized into the necessity for centralization within the vertical structure of the firm (however big it is – from a small business to a national system of government), they still do not account for elements that go beyond the mechanistic aspects of economic relationships and lie in the social plane, such as trust and goodwill. Skin in the game is really needed for decentralized systems to work, but this skin should be very close to one’s body and very private and unique – to the point of scarcity.
Lastly, reflecting on the comment of another panelist (Jungpil Hahn) about the lack of a mere definition of the term “decentralization,” I suggest that its definition depends on the layer – or better, the dimension of decentralization – we apply it to: decentralization of decision-making power is the distribution of ownership (block builders) or intellectual contribution (developers). Decentralization of value capture is the distribution of added value created in the system according to the contribution of private resources (users layer). Importantly, in the absence of a centralized party responsible for the pitfalls of the system, including conflict resolution, litigation, compensation, and fraud risk, the decentralization of responsibility is non-existent – these functions simply vanish away – which affects the perception of decentralized systems. Maybe, the answer is localized decentralized systems with gated control – aka collectives – which, however, do not necessarily need blockchain to operate on decentralized principles.
Arrow, K. J. (1950). A difficulty in the concept of social welfare. Journal of Political Economy, 58(4), 328–346.
Coase, R. H. (1937). The nature of the firm. Economica, 4(16), 386–405.
Grossman, S. J., & Hart, O. D. (1986). The costs and benefits of ownership: A theory of vertical and lateral integration. Journal of Political Economy, 94(4), 691–719.
Hart, O., & Moore, J. (1990). Property rights and the nature of the firm. Journal of Political Economy, 98(6), 1119–1158.
Petryk, M., Müller-Bloch, C., Hahn, J., Halaburda, H., Henfridsson, O., Obermeier, D., & Yoo, Y. (2025). Promises and perils of decentralization in the blockchain age [Panel]. In Proceedings of the 46th International Conference on Information Systems (ICIS 2025) (Paper No. ICIS2025-1702). Association for Information Systems.
Williamson, O. E. (1991). Comparative economic organization: The analysis of discrete structural alternatives. Administrative Science Quarterly, 36(2), 269–296.
Please cite this article as:
Petryk, M. (2025, December 29). Decentralization and the Theory of the Firm. MariiaPetryk.com. https://www.mariiapetryk.com/blog/post-28