by Sasu Ristimäki
From an economic, as well as social perspective, the impact of the internet has largely been a story of disintermediation, which has paralleled by a substantial reallocation of economic wealth. However, the changes in economic value creation and economic organization that underpin the disintermediation and reallocation are still not well discussed or understood.
The commercial age started with Web 1.0, the invention of the web-browser by Netscape enabling the ubiquitous adoption of the Internet. The magic of the ‘Web’ was that any node could directly communicate with any other node, without resorting to any intermediary in between. Economically and socially this was about the disintermediation of agencies, publishers, information aggregators and off-line information intermediaries.
Web 2.0 emerged in the post dot-com years and was defined by the ascendancy of Google in the aftermath of the dot-com crash. The first of the Web 2.0 principles elaborated by O’Reilly et al. was of “The Web as platform”, though key consumer themes were long-tail economics and the power of pricing at zero. Web 2.0 stands out for the disintermediation of bundlers and choice makers and particularly the disintermediation of choice and consumption from inventory. The slow death of the off-line retail industry is still a direct consequence of this.
Web 2.0 gradually morphed into the Platform Economy we are living in now . This has been about disintermediation of end-to-end value chains into value networks, the disintermediation of asset ownership from asset use, and of separating value creation from the ownership or control of assets. Socially, it has been about political organization without the need for or control of political structures, in other words the disintermediation of the political process.
Never before have single networks, let alone privately owned such, been able to connect more than a billion people. And never before have we seen the geometric scaling of network value on such a scale as now, leading to monopolistic information markets and astonishing value capture.
Web 3.0, or the crypto-economy looks to be the age of decentralized systems. The fundamental technological innovation is solving the organization of permissionless peer-to-peer networks , allowing participants in a network to discover, verify and transact with each other on a fully distributed basis. In contrast to Web 1.0 nodes now have the ability to not only communicate but to interact and transact with any other node without resorting to any intermediary in between.
Economically and socially this leads to the potential disintermediation of most centralised parties providing intermediation services, whether they are information intermediaries, transaction intermediaries, or organizational intermediaries .
The likely social and economic impact of this restructuring is hard to overstate, even if it may take a decade or more for the process to unfold.
Economics of disintermediation
As the Internet has evolved we have witnessed a number of industries swept away by a wave we easily term disruption. Individual company failures are described and attributed to any number of specific reasons. But we continue to lack proper tools for the ex ante modelling of changes in value accretion that result from the new forms of economic organization.
Economic theory, and the practical constructs that follow from it, continues to focus on the control of resources and the management of production, even as disintermediation moves economic value accretion further and further away from these aspects.
This results in somewhat of a failure to deal with constructs such as platforms, let alone more complex distributed network systems.
The core understanding of the still prevailing paradigm is that a firm is defined by its assets, and that economic value is generated from the assets of the company in a process of production. Over the past two decades, this process of production has been optimized into a supply chain and much work, both practical and theoretical has gone into optimizing this chain (e.g. Supply Chain Management and ERP). The value is encapsulated in products, which are the full representation of both production value and anticipated customer expectations.
The products are then transacted with customers across a distinct layer, with the firm going to great lengths to convince customers they have anticipated needs and expectations correctly (a.k.a. sales & marketing and CRM).
As the financial representation of the firm assumes first that Equity = Assets and Assets drive value creation, there is a natural conclusion that maximizing value to shareholders (RoE) is equivalent to maximizing asset efficiency (RoNA, RoCE). Despite the intuitive appeal of this representation it is unfortunately diminishingly accurate or useful in practice.
The asset based model is a push model of economic interaction that is unable to grasp generativity (also known as innovation), but moreover is unable to deliver contextuality to the interactions with customers.
The change that needs to be not only recognized but internalized is that value creation is no longer primarily asset based. Actions based on leveraging contextual interactions within multi-party networks are far more productive than asset based transactions. The value creating focus shifts from maximizing asset use, to maximizing interactions and their contextual relevance.
The new model fundamentally drives a change away from closed, end-to-end value chains, to instead dynamically integrate contributions from an entire network around a series of contextually co-ordinated interactions.
The organizational structure that has evolved to capture this logic is the platform: Platforms disintermediate asset ownership from asset use and consequently separate value creation from the control of assets. Furthermore, there may be multiple contributors to the value consumed by the customer. The organizational structure of the platform resolves the multilateral coordination challenges that previously would have been unmanageable.
The ability to connect millions of consumers with millions of providers represents a fundamental change in the information transaction costs. Platforms do not operate without assets, but the assets no longer underpin the value generation. And as the value creation becomes detached from the assets, asset ownership can be disintermediated and detached from consumption.
A platform is essentially a multi-sided information exchange that makes money from resolving transaction information costs. The coordination services provided by platforms allows them to capture up to 20% of transaction value. But, we have no comprehensive explanation for this number and what is it a function of, though asset specificity, contract specificity, network reach, network topology are all likely parameters.
Resolving the transaction information costs of the value chain in a new manner not only re-allocates the value previously captured by intermediaries, but it also reallocates aggregate value across the entire value chain or the new value network.
Despite the work of Nobel laureates Ronald Coase and Oliver Williamson, the theoretical understanding of transaction information costs remains poor. Transaction cost theory still focuses on the question of vertical integration within assumed contractual frameworks, and with optimal control of assets the key parameter to solve for. But the control argument has been made redundant by disintermediation of asset ownership and asset use.
In a distributed system the control of the platform itself as well as the data that underpins the platform are decentralized. The challenge is that as we extend disintermediation even further, the limitations of our theoretical basis become ever more apparent.
Transaction costs and the decentralized economy
The economic motivation of decentralization is that within a decentralized transaction environment there is further ability to capture and reallocate the value created by intermediaries resolving coordination – whether those intermediaries are platforms or legacy structures. However, it is likely that there will also be a reallocation of the other value generation in the chain/network that will take place concurrently.
It is by no means certain that the aggregate value of the new system will equal previous value. In fact there is potential for significant macro deflationary pressure. And, it is not currently obvious how the value distribution of the new system will look like.
Key technological developments that are enabling decentralization beyond platforms are the confluence of the fall in information costs to the point that a non-hierarchical distributed network becomes economically possible; the resolution of the information propagation and consensus problem in unbounded distributed networks (a.k.a. the consensus protocol); the implementation of turing-complete algorithmic logic in non-hierarchical and non-centralized networks (a.k.a. state engines or smart contracts). Add to these revolutionary developments in cryptography that allow the creation of autonomous digital identities and you have the core elements of the blockchain and crypto revolution.
Much attention has been focused on the processing throughput superiority of intermediated hierarchical networks vs. distributed decentralized ones. However, economically the more relevant question relates to the balance of processing costs vs. coordination costs, which is a context related problem. Moreover, the question is typically proposed as a cost problem, whereas it should be presented as a value creation problem. Understanding the difference between the two is critical.
A caveat on institutions and functions
A key caveat, though, relates to a principal mistake in the debate about centralization, which is the tendency to confuse institutions and functions. The Stock Exchange is a centralized institution that provides functions to a decentralized market. Eliminating the institution does not mean that the functions are redundant. To state the obvious, a decentralized system needs to replicate the coordination services of any displaced intermediaries.
This is not a trivial proposition, centralized hierarchical structures with human input are very good at handling high transaction volumes combined with specificity in ways that decentralized algorithmic structures are not. Moreover, hierarchical organizations have a structural ability for parallel processing of information which contributes to their performance.
A challenge particularly of the crypto-movement is that there is often a focus on delegitimizing existing institutions before there is any serious thought of replacement. Furthermore, this is not only a question that is it possible to generate the function of the centralized institution on a distributed basis but is it always desirable to do so.
The libertarian view has been that eliminating centralized institutions is ad deus desirable, and moreover the question of desirability should be a matter for only the immediate participants (with specific priority given to equity holders). We now have enough evidence of algorithmic dysfunction which has arisen from techno- and financial libertarianism that it would seem imperative that the desirability question needs to be considered by society for what is socially desirable. Terms such as crypto-anarchy are quaint in a school yard context but anarchical society is not an objective we should entertain even in jest.
Just as from an economic perspective we should not underestimate the potential of decentralization to lead to disruption, we should not do so from a social perspective either. In fact, it is important to understand how strongly the two are likely to be interlinked. But whereas economic disruption can be socially beneficial, social disruption is somewhat more problematic. Views on the desirability of social disruption differ depending on the social context. In jurisdictions where laws are not enforced, or worse are enforced selectively and arbitrarily, the push to disrupt and disintermediate institutions is understandable as these institutions have little or no legitimacy in the first place.
Introducing digital currency
One of the major conceptual gaps to be bridged remains the structuring of the new decentralized systems. While platforms disintermediate asset control focused structures, they are still centralized structures in terms of information control, governance and value accretion. In short, they are still classic corporations in many respects. The ultimate objective of a distributed economic system is to decentralize information control, decentralize governance and also redistribute value accretion.
It is not certain that the resulting structure will fit our definition of corporation any longer. But the new practical challenge is how to structure the operating logic and information exchange within the new decentralized structure. This is where digital currencies or tokens come into the picture.
The concept of digital currency has been first implemented (in scale) in the Bitcoin protocol as the vital new element to resolve the problem of propagation and consensus in distributed, open, peer-to-peer networks. Achieving consensus in such a network was a problem identified already some decades ago and which for a long time was considered to be unsolvable. The solution provided by Bitcoin is a combination of a reward mechanism, combined with a proof-of-work a mechanism, to incentivize and select nodes, that work to validate new transactions and update the consensus of the network.
The validation service that the reward mechanism underpins is clearly valuable for the network members, which is why the reward mechanism is successful in transferring some of this value to the validating nodes as an incentive for their work. The core of the reward mechanism is bitcoin, the token which is now termed digital or crypto currency.
The token is generally accepted within the network as a means of settling transactions and transferring value between network members. It is valuable to network members because of the value of the information propagation and consensus it underpins. And, membership of the network is open and voluntary. Anyone can choose to be a member of the network, but no-one is compelled to be a member.
Thus, within the scope of the network, the new digital currency acquires characteristics akin to those of money. There are important caveats, though. Firstly, those characteristics do not extend beyond the network. Whereas within the network the crypto token may be valuable, the value proposition outside the network is tenuous. Secondly, whereas the crypto token is valuable because of the consensus it underpins for the network, the magnitude of the value proposition is not simple to model. The value of the token is clearly related to the value of the network, but beyond this high level of abstraction there is significant unresolved debate on the details, and underlying value may often be (much) less than nominally made out be .
Digital currency as an information mechanism
Even as general attention has focused on the financial properties of crypto currency (financial speculation invariably seems to accompany all new ideas) we should re-focus on the original core property which is to resolve the consensus problem of distributed networks.
As the networks have evolved to support more complex functionality it has been the token which has also grown in significance as the principal mechanism for not only resolving who owns what, but on communicating and intermediating increasingly complex informational and operational content.
In essence, the digital currency is the principal mechanism for resolving transaction information costs within the distributed structure, that itself is all about resolving coordination on a decentralized basis. To simplify, a protocol is the procedural set or system of rules that governs a network. The token encapsulates the protocol.
Bending our thinking
We understand companies structured around equity; we understand centralized organizations that typically operate as hierarchies. Our theories of accounting, finance, management and corporate economics are based on this understanding.
It is not simple to adapt so much accumulated thought. Just as happened as platforms emerged, it is all too easy to dismiss the significance of these new structures out of hand. Only in 2012 many blue-chip executives were still questioning the relevance of Facebook, while Uber which is now eyeing their own $100bn+ IPO was dismissed as just a ‘taxi-app’.
Yet the same underlying forces that have driven the Internet disintermediation trend so far are inexorably pushing us towards decentralization. Economic value creation no longer revolves around the question of how to most efficiently resolve production costs. Platforms have established that maximizing value creation requires instead focusing on resolving transaction information costs asking how to most efficiently maximize coordination ability. Once we change the questions we ask we are likely to end up with a new appreciation of the new answers.
References and notes
The term Web 2.0 was first coined by Tim O’Reilly and Dale Dougherty in 2004 https://www.oreilly.com/pub/a/web2/archive/what-is-web-20.html?page=1
It could be argued that the Platform Economy is already a new web, Web 3.0
Though disintermediation is possible in principle, it is not a trivial challenge to execute or implement this in practice.
For a more comprehensive discussion of how the enterprise paradigm is changing see https://startupsventurecapital.com/new-metrics-for-the-algorithmic-enterprise-2192d4402651
For a discussion on hierarchical structures see Mengistu et al., The Evolutionary Origins of Hierarchy. http://journals.plos.org/ploscompbiol/article/file?id=10.1371/journal.pcbi.1004829&type=printable