The Economic Value of Decentralized Identity — Part 2

Reimagining the economics of trust and reputation

[ Note: this is a version edited down from a longer whitepaper titled “Decentralized Identity as a Meta-platform: How Cooperation Beats Aggregation,” published by the Rebooting the Web of Trust biannual conference, available here in its canonical form. It was written collaboratively by Michael Shea, M. Smith Samuel M. Smith Samuel M. Smith Ph.D., Carsten Stöcker Ph.D., with editing and contributions from Juan Caballero Ph.D. and Matt G. Condon. ]

In the first installment in this series, subtitled “The theory of the ‘meta-platform’ network effects and IT power-plays,”we sketched out an abstract theory of “the meta-platform,” trying to think beyond the orthodoxy of network effects by interrogating their competitive presumptions and zero-sum mental models. Chipping away in particular at the predominance of Metcalfe’s Law as the guiding light of internet business generally, we proposed a set of conceptual tools for thinking about cooperative and interconnected platforms. Having set out these building blocks, we now turn our attention to the complex nature of trust in such an interconnected and cooperative landscape.

Accounting for trust in current and future systems

A common approach to understanding business processes today is via the model of transaction costs. In this tradition, the lion’s share of business value is seen to come directly from a net reduction in transaction costs for all participants in a given value chain. Transaction costs may be grouped into three classes:

1. triangulation,

2. transfer,

3. and trust.

Triangulation and transfer costs are relatively uncontroversial here, and do not display fundamental changes between tradition and more-open platforms, so we will not discuss them here at length.

Trust, on the other hand, may well be the most important of the three for platforms and meta-platforms. What’s more, it may be the most affected by the closed/open distinction. Without trust, platforms, or any other kind of sophisticated physical and digital value chains, cannot develop to maturity. Trust and reputation are truly paramount to commerce as we know it; they are the currency of brands and the building blocks of marketplaces.

The “trust tax” stabilizing and structuring value chains today

The requisite level of trust for most commercial and business processes has been historically difficult, and costly, to establish. Trusted suppliers are thoughtfully selected and managed, as well as monitored and certified for quality, reliability and consistency. Regulators demand certifications and audits to ensure that best practices are followed, and companies such as TÜV, Underwriters Laboratories, and Intertek conduct inspections and provide quality certifications.

At the end of the value chain, customers purchase products because they trust the quality of the brands, which is the key differentiator for many companies. The difference in price between a widely known and a lesser-known brand, or between a national brand and a more local brand, or between branded and generic products, illustrates how at every point in the chain, higher levels of trust come at a notable premium.

This premium, sometimes referred to as a “trust tax,” that consumers pay to trusted global brands is levied all the way upstream, throughout the brands’ supply chains to their suppliers. From the use of “conflict-free” raw materials, to Six Sigma manufacturing practices, to Independent Standards Organization (ISO) certifications, to the validity of shipping or purchase orders — the work of verifying all these characteristics “entity to entity” requires endless chain-letters of paperwork, endless e-mails and phone calls, and costly site visits and audits. These cumbersome activities reduce productivity and efficiency throughout the economy, costs which are ultimately passed down to consumers.

The immense volumes of data generated by global supply chains has other effects as well. If such data were properly mined and analyzed in a verifiable way, the resulting analysis would help to establish the trustworthiness of an entire value chain process and its resulting products. However, assembling and cleaning the requisite data set for such an analytical process requires numerous manual interventions and the involvement of many parties across multiple platforms.

Furthermore, these kinds of audits are usually done by neutral consortia or by outside firms whose incentives do not touch the competition in the relevant markets, either of which could feasibly centralize data access across a market without triggering competitive defenses. Competition for control of such consortia, or distrust of such profit-driven centralized firms, can slow down the integration of such audits, which adds yet more costs. In few cases is this kind of deep business intelligence analysis practicable.

Identity and reputation systems as trust brokerages

What makes the kind of end-to-end analytics described above so rare is that too much of the requisite data is hoarded and guarded in siloes, without much concern for the best interest of a healthy trust market, or for that matter any other market-wide concern.

Trust is rarely studied at market scale, and relevant data sets for doing so have been hard to amass. It is our belief that if such a bird’s-eye perspective could be attained, there would be many net benefits to empowering individual data subjects to hold and reuse more of the data on which their reputation is calculated. Doing so would allow “credential-based” (or, to use a buzzword du jour, “data-driven”) trust among participants that has never been supported before.

But doing so requires loosening the grip most identity systems currently have on the credentials they already use (internally) to gage and estimate the trustworthiness of their participants, i.e., to calculate “reputations” for all participants in their networks. The same credentials that identity systems use to do this, if handed over to the participants, would allow them to prove their trustworthiness to other participants of their own choosing or on other platforms. In this way, the raw material of reputation systems would take on a different kind of value, circulating more widely and being reused again and again anywhere a given subject can prove the validity of those credentials,i.e., anywhere they can prove their own identity.

Traditionally, identity and reputation are seen as interdependent functions of one system, or even as two sides of one coin. After all, a reputation is meaningless without an underlying identity and an identity is valueless without a credible reputation. After a few decades of increasingly centralized digital identity systems, we have come to take it for granted that identity companies should naturally keep a close grip on reputation services which they are especially positioned to provide for others. Indeed, the colloquial terminology can even be an impediment to grasping this division of labor where credentials and tracking add up to valuable user analytics sold to the highest bidder. Because identity systems typically include credentials of some form that encode reputation as data, the software industry usually refers to credentialed identity systems simply as “identity” systems, not as “identity and reputation” systems. When the calculation of reputations becomes primarily behavior-based rather than credential-based, these systems are typically referred to as “reputation systems” rather than identity systems, even though they are still dependent on an underlying identity system and its credentials.

For our purposes here, we view all of these traditional identity and reputation systems as existing on a spectrum of “trust conveyance”. While making credentials more portable across organizational boundaries can incur its own costs and complications, it has a curious network-of-networks effect whereby all parties benefit from more information and more possible because trustworthy connections. Maximizing trust conveyance works out best for everyone except for today’s monopolists of trust.

Most importantly, a decentralized identity and reputation system enables trust to be transitive between contexts to some degree. This additional dimension of cooperative network-of-network effects brings trust transaction costs down generally. This “knock-on effect” can be quantified by adding a transitivity factor to the equations above.

What’s more, organizing trust in a conveyance-maximizing way for individual participants also allows for the trust or credibility of all data for auditors and platforms. Situations where trust data is maximally portable enables end-to-end (E2E) verifiable audit trails along a given value chain even if it crosses many organizational boundaries. “Maximally portable” might sound like a high bar or systemwide buy-in, but it is readily attainable anywhere (and for as long as) all parties can be incentivized to minimally uphold data and control interoperability.

Imagining a trust landscape where reputation data flows more freely

A system-wide trust mechanism along these lines can be seen at work in the decentralized identifiers and verifiable credentials of the sort standardized by the Worldwide Web Consortium (W3C), in combination with consequently interoperable agents, interaction protocols, and credential-storing wallets. We call this a “meta-platform”, because it enables participants to engage in trusted interactions in a wider ecosystem and maximizes intra-network value transfers.

This approach can slash the “trust tax” at every level, starting with participants agreeing to implement and follow the existing decentralized identity standards and their future developments. This decentralized approach does not build another aggregator platform or another proprietary blockchain solution, but rather a participant-controlled decentralized meta-platform. Opinions vary as to what this participant control could or likely would look like, but some proponents of decentralization argue that bottom-up governance is easier to institute in these kinds of meta-networks.

Because the primary value of a platform is to reduce transactions costs for all parties, a decentralized identity system makes the building-blocks of reputation more portable. Even if changing nothing else, this fundamental change at the root of trust transactions could significantly reduce average, net, and/or aggregate trust transaction costs in many different marketplaces and other industrial contexts.

This may make whole families of transactions viable that were not before, or create new kinds of commerce (and perhaps even self-regulating ones). This increases the value of the associated platforms per participant and lowers the critical platform size (see below). This further accelerates network-of-network effects. One quick example should make this more concrete.

On-Boarding: A metaphor for network participation

One early application of Metcalfe’s law in the business models of the ascendant communications industries was to show that even large upfront network connection costs would eventually be overcome by the exponential increase in value due to network size. The so-called “break-even point” was the point in time where the growing network’s value to a participant overcame (due to network effects) the cost of joining the network. The size a network has reached at this point is called the “critical network size”.

We could adapt this terminology to today’s platform economics by pointing out that the major upfront cost of connecting to a platform is not the internet connection itself but the time an end-user spends creating an account with login credentials and provisioning electronic payment, with concomitant identity verification needs. In the traditional “scale”-business models focused on the apps and products of Web2, this is the central force of inertia to be overcome.

Unbeknownst to most end-users, minimizing this time cost and thus overcoming this inertia typically requires the participation of many companies and market-wide mechanisms (from insurance to credit cards to underwriters to regulatory bodies to infrastructure providers), all of whose costs are baked into our current networks of global commerce. But also baked into this system is a unified user experience of convenience and relative simplicity. In the case of mobile phone app stores, the unification of user experience is a powerful market effect of centralization.

One of the problems with decentralized blockchain technology is that in general it still needs to make substantial progress towards smooth operation and user experience before it can compete with this status quo. In terms of “on-boarding” a user to a new platform or network, many radically peer-to-peer digital networks or blockchain systems have quite high on-boarding costs. In the case of peer-to-peer technologies, there can be a very steep learning curve and “UX deficit” relative to increasingly convenient and intuitive commercial software.

In the case of blockchain and cryptocurrencies networks, participants have to contend with difficulty in managing keys, increased regulatory friction, and a level of complexity quite high by contemporary standards. The plethora of competing (and thus, at least to date, largely non-cooperative) blockchain platforms only heightens the confusion. The result is that critical platform size may be significantly increased, with the worrisome effect that many blockchain platforms may never reach their critical size (or even a “break-even point” at which their scale and overhead are practical for their intended use cases).

A decentralized identity meta-platform allows those on-boarding costs to be amortized across every platform a participant chooses to join. This potentially lowers the critical platform size (and the break-even point) for the participant on each of the sub-platforms. None of this directly helps to unify user experience, but perhaps on the whole this “transitive onboarding” can offer some incentives towards cooperation between platforms, in design terms at least. All of this could readily accelerate network-of-network effects and overcome some of the inefficiencies created by competition.

Applying this same analysis to more complex and industry-specific examples, as we will do in the next installment in this series, should make clear that this is not an abstract or academic theory but a major shift in the foundations of economic practice. Sooner or later, we believe, it will come to every industry, but we will focus on those where we have the most insight.