By Arik Ben-Zvi and Steve Weber
A shortened version of this piece was originally published in Noema Magazine.
When Karl Polanyi wrote The Great Transformation in 1944 he could have hardly imagined that — nearly 80 years later — the story of Web 3 would repeat nearly verbatim the narrative of state-society-market interdependence that he laid out for an earlier era.
An economic historian and sociologist born in Vienna, Polayni lived through experiments in Marxism that tried to reorganize society and economy on the basis of a radical theory that placed technology and economics ahead of society and culture as the driving force of history. Rejecting this world-view, he created the concept of a ‘double movement’ where efforts to dis-embed economy and markets from society were bound to fail — again and again. In his era, it was laissez-faire economic ideology that argued for putting markets ahead of all other forces in order to create a market-driven society that could overcome the felt stagnation of the time. He saw it as inevitable that society would reassert itself and work to re-embed the economy against the commodification of all things, demanding limits like trade protection and labor laws that a purely market-driven system would abhor.
This important insight had a simple and profound implication: the notion of a self-organizing and self-regulating market was and would remain utopian (or dystopian, depending on your perspective). Markets depend on governments and social forces just as much as the other way around. And though Polanyi wouldn’t have recognized the cryptography protocols that power Web 3, it’s certain that he would have well recognized the utopian ideologies that accompany those protocols, as simply the latest attempt by human beings to escape what Marx called ‘the fetters of social relations on the means of production’.
The problem is, they aren’t just fetters and they can’t be escaped. Not just ‘shouldn’t’ — can’t. In this essay, we explain why and more importantly, what Web 3 entrepreneurs can and should do with that understanding in hand.
It wasn’t that long ago when…
The communities of the 1970s and 80s that famously gave birth to the personal computer and the first phase of the Web (Web 1) had deep roots in ideologies drawn from Northern California libertarianism as it developed alongside the Grateful Dead, the Merry Pranksters, the Whole Earth Catalog and the early environmental movement. This story has been told many times in pieces (and is about to be told most eloquently in John Markoff’s forthcoming biography of Stewart Brand, an iconic figure in each of these moments). What’s less well understood and remembered is the importance of Ronald Coase as the unlikely intellectual muse for Web 1 and the dot com era. The Coase theorem became the herald of a new and profound laissez faire ideology that didn’t just argue (as did the Reagan-Thatcher neoliberal movement) for the reduction of government control of markets. It went further to propose that a society built on marketizing literally every possible social interaction would be more efficient, more sustainable, more fair, and more gratifying and meaningful for human beings.
The Coase theorem in its original articulation (The Problem of Social Cost, 1960) was a simple thought experiment that goes like this: if transaction costs are sufficiently low and property rights clear, then the initial allocation of resources at any moment doesn’t matter because markets will trade things around until each resource lands in the hands of the owner who can put that resource to its optimal use. In Coase’s world there really are no meaningful politics of allocation or distribution; political distortions and legacies simply get traded away by markets.
And so the digital idealists of this era extolled the internet and the web in particular as a transaction-cost reduction system first and foremost, enabling the world to move toward a Coasian ideal. Along with that view came globalization, the reduction of political and geographic barriers to market transactions around the world. And the wisdom of crowds, the reduction of decision rights in the hands of so-called experts. And reputation as the new currency of power. It was a heady time, with manifestoes that celebrated the liberation of human potential through and in markets and the digital world as a politics and policy-free zone that would float freely about the tawdry layer of politicized institutions, gatekeepers, and legacy arbiters of truth. All of this was to be replaced by the radical democratization of culture, creativity and epistemology through real time auctions with zero barriers to entry.
The Cluetrain Manifesto (a set of 99 theses posted online in 1999) crystallized Web 1 thinking in the tagline ‘markets are conversations’. What this implied was that somehow demand and supply — the conversing parties — were to be equally powerful in this new and apolitical era. It obviously didn’t turn out that way. And we now know the obvious and entirely foreseeable reasons why.
Network effects led to new forms of self-reinforcing market power. The early leaders turned out to have a persistent advantage in markets even if they were no where close to optimal. This in turn led to political power as the early winners in technology standards and platform businesses discovered how to use their wealth and popularity to control or overwhelm the political and social counterforces that raised questions and sought to constrain them. In plain English, the Coasian utopia devolved into something more like a plutocracy, where the rich find it within their capability to control the rules that shape markets in ways that enhance and reinforce their asymmetric ability to amass yet more wealth and power. The world’s largest corporations became the world’s most powerful influencers of political actors, and thus the shapers of non-market rules.
Fast forward 20 years, and it’s a picture we know all too well in what is now called Web 2. The ambition of today’s would-be political reformers can now be felt at almost every level of government, from the global to the continental to the national and state levels. Polanyi’s double movement is having its modern moment as they begin to chip away at the accumulated plutocratic power of the world’s largest technology platforms through anti-trust and a host of other regulatory initiatives aimed at re-asserting political and social control. The precise outcomes are uncertain, but the reaction is now in 2022 in full force, from Beijing to Brussels and Washington (ironically it may be strongest at the moment in Beijing). And of course in public opinion.
And in the midst of this comes another reaction, a kind of countermovement to the slow and frustrating progress of the double movement itself. This is the ideology of Web 3 — radical digitization, decentralization, and democratization that seeks to replace governance with consensus and formal organizations with DAOs. The underlying themes are entirely familiar to anyone who lived through the 1990s — as has now been pointed out by many observers.
But there is an important difference as well. The radical marketizers of Web 1 didn’t spend much time thinking about what representations of value would serve to denominate and settle transactions in their Coasian paradise. But Web 3 struck first and directly at one of the most central elements of what it means to be a sovereign political entity, control over money.
It’s no coincidence that DAOs, metaverses, and other Web 3 concepts rest on the use of cryptocurrencies as their medium of exchange, store of value, and unit of account. Yet there’s no law of nature operating that says it has to be this way: in principle, there is no reason why a DAO couldn’t use dollars or any other fiat currency and exactly the same is true of an immersive 3 D computing environment (you could certainly buy a skin for your avatar with euros just as easily as you can do it with bitcoin). But Web 3 isn’t just a new set of technologies or business models, and it isn’t just the substitution of blockchain for central ledgers or spreadsheets as a matter of efficiency or convenience. Quite the opposite — for most applications, using blockchain right now is less efficient, less convenient, and far less user-friendly than conventional alternatives.
Crypto currency and the blockchain are central because they represent an explicit agenda to route around and eventually undermine centralized authority, whether it be formal governance institutions, traditional corporations, or central banks.
Web 3 champions certainly have a point when they argue that each of these ‘legacy institutions’ have over time become calcified and inefficient, unnecessarily expensive, inadequate in their representation of diverse human interests, and in some cases outrightly corrupt. Subjecting these institutions to greater competition (many have become de facto monopolies) and forcing reform and improvements by doing so can only be to the good. Web 3 really does have significant potential to bring some greater balance to the power and wealth asymmetries in modern life (for example, a crypto-powered global remittance system would at a minimum drive exploitative money transfer organizations to reduce their excessive overhead take and transaction fees, which hurt some of the most vulnerable people on the planet).
But that competition and reform narrative isn’t where Web 3 ideology stops right now. The stated ambitions are much greater. Web 3 wants to go all the way — by ignoring, routing around, and ultimately driving into irrelevance the non-market forces and actors that get in the way. It’s another utopian (or dystopian) market-world vision, and it too will get caught in Polanyi’s double movement. The next section explains why.
Non-Market Foundations for Web 3
The core proposition behind the distributed ledger technology that drives Web 3 is simply that transparency and consensus can better establish trust and confidence in essential ground truths of social interaction, than can centralized institutions. The argument is sometimes made in abstract philosophical terms, and sometimes made by pointing to the many failures of centralized institutions to maintain trust and confidence over time. So what do we actually know from experience about this crucial proposition? In fact, we know that confidence and trust is a non-market phenomenon at its core, because it is to non-market institutions and forces that people turn when the going gets tough.
The typical equity market is a case in point. The blockchain revolution of the last decade has shown that it is possible to establish a meaningful level of trust and confidence in these markets via software, DLT, and consensus mechanisms — during normal times when markets are stable. Those times are not, however, the acid test of confidence and trust. When markets act as expected, with familiar price fluctuations and the like, the burden of proof is light. It’s when markets come under extreme stress that confidence and trust is put to the test. And while a decade or more can sometimes pass without the extreme and unexpected events that roil markets, these events and disruptions are episodically inevitable simply because of the nature of complex markets and the psychology of risk. Hyman Minsky called this ‘the instability of stability’, like adding more sand to the top of a sandpile that appears entirely stable and thus invites you to add more sand — up until the moment that the avalanche starts. The analogy is to stable markets that lead participants to take on increasing risk — which don’t seem that risky at the moment — until the inflection point inevitably arrives.
The pertinent question then becomes, can private money based on DLT and consensus mechanisms retain confidence during periods of severe financial market stress? The multi-century history of market panics gives no cause for optimism here. Consider the most recent example of the 2008-9 financial crisis and CDOs (collateralized debt obligations). These exotic instruments — the product of the best and most innovative minds in the industry operating under a permissive regulatory umbrella and private rating agencies — were attractively stable, profitable and seemingly efficient at repackaging risk. Until the moment that they weren’t, which led not to ‘normal’ price adjustments but to crisis, collapse, impending market failure, and the necessary intervention of centralized institutions acting as the lender and buyer of last resort to forestall what would have otherwise become a global great depression.
Who today would serve this rescue function for crypto markets? Regulators are right to pose this question, and they have a point in taking the position that being dragged into a must-rescue mode without any prior influence on how markets evolve creates massive moral hazard that is good for no one. The more general point is that confidence and trust are non-market phenomena when it matters most, which is when markets are under stress. And even if the DLT champions mount theoretical arguments to claim that it need not be so in some abstract sense, this position of caution will — reasonably given the history — remain the default position of most regulatory agencies. Not necessarily because they are self-interested, rent-seeking, or just resistant to change, but because they don’t want to make the same mistake yet again.
In this context money in and of itself is a non-market phenomenon, or it might be even better to say a ‘pre-market’ phenomenon. Despite the persistent gold-bug ideology that wishes it were otherwise, extreme versions of apolitical money simply do not work. A bounded and completely inflexible system of money supply (ie if there were only bitcoin, and only 21 million can ever be mined) is inherently deflationary, and would restrain and distort economic activity as badly or worse than does inflation. At the other extreme, it should be self-evident that completely unrestricted money — money that anyone can create at will — would immediately lose any ability to act as a store of value, a meaningful medium of exchange, or a unit of account. Every form of money that has stood the test of time lies somewhere in the middle.
And so consider the example of programmable money — that is, tokens that carry with them instructions or restrictions and are not fully anonymous or fungible — which can be traced, restricted to certain uses, or modified in time (such as with negative interest rates). All of these characteristics are about shaping market interactions by favoring some transactions over others (including on the dimension of time horizon — a negative interest rate makes the same transaction more attractive today than it would be tomorrow). In fact, the whole point of programmable money is to create incentives and commitment mechanisms that go beyond what markets would do in and of themselves. The person who wants to bind herself to save 20% of her salary for retirement and uses programmable money to do that is in practice using a token to prevent herself from succumbing to short term temptation, that is hard to do given human inclinations toward short term bias (otherwise known as time inconsistency of preferences). The person how choose to receive her salary in ‘green money’ that can only be used for exchange between environmentally-friendly companies is binding herself in a similar way. This is much more than reducing friction, transaction costs, and time to settlement — since all of those blockchain functions are market-improving or market-perfecting; programmable money is almost the perfect tool for non-market interactions that then shape the market.
And of course laws themselves are a non-market phenomenon. DAO enthusiasts are quick to point out that discrete decisions can indeed be made within market structures. But what about the familiar question that comes before the decision — the decision rule, or voting system if that’s the way decisions are made? How do markets write their own decision rules? The answer is, they don’t, and in a fundamental sense, they can’t. It’s not a coincidence, then, that the DAO community is celebrating the fact that DAOs were granted legal status in Wyoming. A DAO didn’t grant that status; a market-shaping legislature did. Property rights in DAOs, as in most modern settings, ultimately depend on non-market interactions if you look sufficiently hard behind the curtain.
No Blank Slate
It’s a simple political reality that social movements, even those based in revolutionary technologies, don’t get to wipe the slate of history clean and start over. Autos had to navigate roads that were built for horse carts, just as the first generations of internet protocol data had to travel over copper telephone wires. Change comes, but it’s a more gradual and negotiated political process. It’s surely frustrating to evangelists but still the case that today’s institutions like banks, firms, courts, and government bureaucracies aren’t going to simply step back and let themselves be overtaken by DLTs, DAOs, and radical democratization. Incumbents will resist in both legitimate and sometimes illegitimate ways, and ultimately negotiate a division of labor between the new and the old. We can already see three pathways that this negotiation will take.
There’s a category of incumbents for whom Web 3 is an existential threat, because they are pure middle people who collect rents by layering transaction costs and friction on markets without adding meaningful value. Everybody encounters these incumbents over the course of their lives and typically wonders why on earth do these institutions exist? Usually the answer lies in history — they are legacies of inefficiencies and information asymmetries that made some sense at an earlier time but just have no logical place anymore in today’s world. Consider for example the substantial fees you pay to a ‘title insurance’ company and ‘escrow services’ when you buy a house, or the requirement (still present in some US states) that manufacturers cannot sell cars directly to consumers but must have physical dealerships present in the customer’s state. These incumbents will put up a nasty fight with Web 3 – they are well practiced at protecting themselves and they have nothing to lose by doing so. But they won’t win in the long term because they have no real ground to stand on. Web 3 will push this category of incumbents out of existence. Just don’t expect it to happen quickly or without real pain.
There’s a second category of incumbents that can and do provide real value, but are visibly inefficient, take excessive rents, and are simply more expensive than they ought to be. Many governing and legislative bodies fall in this group — they do work that is important and valuable but often do it poorly. Obvious examples come from institutions that create value through support of ongoing negotiation and compromise among contending interest groups. Smart contracts can implement agreements efficiently, but they can’t negotiate the agreement itself. Algorithms can in principle identify contract zones among static interests, but an important aspect of good negotiation is about more than that — it’s also about upgrading and articulating a common and collective interest that is greater than simply a tradeoff between the predetermined interests of contending parties. For example, good tax policies that support long term social cohesion, investment in future generations, or even high-risk innovation policy that will benefit yet-to-be created companies that can’t speak for themselves in a static negotiation because they don’t yet exist.
Web 3 can drive these kinds of institutions to become more efficient by constraining their rent-seeking behaviors and reducing unreasonable profits. It might fully replace some of them. What’s certain is that it will be a robust competition, and that’s a good thing. For example, blockchain-based payment and settlement systems might or might not fully replace credit card networks and their legacy payment rails, but they will almost certainly drive down excessive exchange fees that the networks have been charging (and defending) for decades.
And there is a third category of institutions essential to the functioning of markets and society that are seemingly impossible to replace with decentralized, consensus-driven alternatives. Courts are the most obvious example. DAO enthusiasts will dispute that claim, but they have yet to explain how smart contracts can manage the existential economic reality of incomplete contracting — that in any reasonably complex real world setting, it is simply impossible to foresee and write down all contingencies that can and do emerge in the course of executing a contract. How would a smart contract have dealt with the multi-billion dollar insurance disagreement over whether the World Trade Center attack constituted one event, or two events? Is it possible to even imagine a comprehensively exhaustive definition of what is an ‘act of war’ that is stable over time? Courts and related adjudication institutions are essential for dispute settlement in ambiguous situations without full information and where judgement calls must be made. Smart contracts fail here because what is in dispute is most often not a matter of fact, but a matter of interpretation — and this is true whether we are talking about spousal divorce suits or competition policy (where the most important consideration is how to define the relevant market in which market concentration is assessed).
Do we really want to assign these kinds of interpretive tasks to decentralized voting or consensus mechanisms? Only if we can imagine having solved the evergreen problem of the tyranny of the majority. Or are willing to bet our collective futures on decisions of the crowd. And removed somehow the ever-present possibility of manipulation and inappropriate persuasion. But imagining that can be done or would somehow naturally emerge in Web 3 words simply pushes the problem back to the naive starting point of Web 1: that people are inherently ‘good’ in ways that others also recognize; that people seek truth and justice rather than personal gain at the expense of others; and that crowds can define what is right on their own. Theoretical possibilities aside, it simply hasn’t worked out that way through history.
There’s one more political reality that is distinctive to this moment in history, but still critically important. Cryptocurrency enthusiasts like to point out that some of the ills of the modern US financial system are certainly a function of the fact that the US financial sector is concentrated like a power law, with a few very large institutions and a large number of very small ones. But it’s not nearly as concentrated at the top as is Big Tech. If Web 3 becomes associated – rightly or wrongly – in the publics’ and regulators’ minds with big tech, then the argument for Web 3 as a decentralizing and democratizing force in the digital world will simply fall apart.
What is to be Done? Building a Non-market Strategy for Web 3
Web 3 deserves the opportunity to explore its potential; there’s too much promise to lose that chance. The goal of this essay was to name the issues that could short-circuit it, and develop an approach to strategy that enables broad experimentation. We put forward in that spirit a brief mindset perspective and a practical to-do list for the near term.
The mindset perspective is a combination of two kinds of radical acceptance. The first is radical acceptance of the Polanyi double movement as ground truth. Non-market forces shape markets; they can do so well or poorly but they won’t be ignored. Web 3 enthusiasts can own this truth and deploy non-market influence to aim for a playing field that is competitive and experimental. The greatest risk to Web 3 isn’t really governments and regulators; it is incumbents that use governments and regulators to avoid competition and particularly the kind that comes from competitors that don’t look or act like them.
The second is radical acceptance of the proposition that trust is always and everywhere a social-psychological phenomenon not purely a material one. And the social psychology component of trust is most needed at moments of maximum material economic and market stress, which will occur repeatedly and inevitably. Overly aggressive narratives that proclaim revolution and ‘this changes everything’ don’t really serve the goals of Web 3 hopefuls. The most effective revolutionaries don’t need to talk about themselves like that. And it isn’t a good way to build the broader foundation of trust that Web 3 will need when things get ugly — when there is a cryptocurrency panic, when a significant security flaw is discovered in an important blockchain protocol, when a major terrorist attack is financed through tokens, or some other unforeseeable but analogous event.
Radical acceptance is a mind state that is anything but passive; the point is to enable courageous action. The first courageous action to take is this simple: drop the old and tired meme that ‘regulators are dumb and just don’t understand technology or what we do, so they need to say out of our way’. Instead, respect their risk assessment and mitigation mission for what it is, and make serious and sincere efforts to educate them and their public constituencies. Regulators should be thought of like directors on boards of publicly traded firms – not charged with technology decisions per se, but with the equivalent of enterprise risk oversight in regard to those decisions. Washington DC wasn’t filled with petroleum engineers when the federal government brought anti-trust suits against Standard Oil, and it didn’t have to be. Today’s blockchain and crypto sector isn’t categorically different in that respect.
This means the Web 3 community should work to assist and support the non-specialist community with tools for high quality risk assessment. Which in turn means full transparency about risks that are known and those that are unknown. For example, it is true (and very important) that crypto’s ability to reduce transaction costs and friction in cross-border payments has many upsides. But it also has foreseeable macroeconomic downside risks, particularly for some of the same emerging markets that are said to benefit from the upside. Here’s one way this works: reducing friction in cross border capital flows almost certainly will make capital flight easier. It is essentially the opposite of a Tobin tax and other partial capital control mechanisms that emerging markets can deploy to prevent volatile, whiplash capital flows (and particularly, the sharp outflows that sometimes occur during economic downturns, or when the Fed raises interest rates and makes US Treasuries a comparatively more attractive investment).
Crypto could very well exacerbate these rapid fluctuations, which bring real macroeconomic pain to emerging markets. Yet it’s rare to find anyone inside the crypto or DeFi enthusiast community talking openly about this risk and how it could be mitigated. It’s an essential part of non-market strategy to look these risks straight in the face and grapple with them voluntarily, rather than being forced to do so by others. Don’t wait for someone else to build the risk assessment dashboard that boards and regulators want — build it yourself. And don’t just assume that more education and understanding of crypto and Web 3 will automatically lead to more positive attitudes about it. What it should lead to, are more thoughtful and sophisticated iterations of the risk dashboard.
A second courageous action would be for Web 3 organizations to take the lead on data protection and safety issues in a way that Web 2 companies generally have not. A simple rule of thumb for Web 3 might be to treat regulation here as a floor not a ceiling. The non-market logic for this is straightforward: the things that went wrong on data protection and safety in Web 2 social media and overall in cybersecurity could be multiplied many times over in a Web 3 immersive and decentralized environment. More important, regulators and many consumers have every reason to believe that this will happen, or to accept at face value the arguments of digital skeptics who tell them it will inevitably happen. It doesn’t take much to imagine how the rhetoric of ‘surveillance capitalism’ gets supercharged in the metaverse, and it’s already in progress.
Consider Section 230, a backbone regulatory enabler for the digital economy that is right now struggling to maintain its tenuous equilibrium hold on the non-market landscape even for Web 2 companies. The architecture of Web 3 constrains the choice set for how to evolve regulation of this sort – how can a central authority be held responsible in any manner for user-generated content, if there is no legal entity that acts as a central authority but only a code base developed and maintained by an open source community of volunteers? This argument as made by DAO proponents has an unassailable internal logic, but it doesn’t win the day. It can’t be an excuse to ignore the very real issues that arise in harmful content that lives on that code base. Or to go backwards and indulge in modern versions of the naive Web 1 ideology that decentralized consensus mechanisms, the wisdom of crowds, and reputation economics will filter out the bad and multiply the good. Web 2 taught us that Gresham’s law — the observation that in an unmanaged setting bad money drives out good money — applies to ideas as well. Web 3 proponents need to face this problem straight on and get ahead of it. No one right now has the precisely right mix of tools — including machine learning systems for content moderation, social norms, outright rules, and others — to know how to do this, which is the best argument for broad, radical, and deep experimentation. Ignore or postpone attention to this, and non-market forces are going to close the window on Web 3 sooner not later.
A final item for the to-do list is a bit more general but crucially strategic. Every non-market move that a Web 3 champion takes under consideration should be put through this simple decision filter: does it make it easier, or harder, for regulators to give us license to experiment. This is where communications matters a great deal. If you proclaim your mission is to undermine the power of the US government, the Federal Reserve, Citibank and Goldman Sachs, Meta and Google, you should fully expect those institutions to move toward overcoming their differences just enough to work together to crush you. If you proclaim your mission is to build alternatives that bring pressure and competition to bear on legacy institutions, you’re more likely to get a chance to prove that you can do it.
Here’s a simple example: stablecoins backed by US dollars most likely won’t and certainly don’t have to undermine the power of the dollar in international finance — so why would anyone talk about it that way? In fact, the impact on the dollar is more likely to be positive (for example, dollar-linked stablecoin tokens would make it easier to gain access to dollar-denominated assets in parts of the world where demand for those assets is high, such as many emerging markets that suffer from deficient governance and low confidence national currencies). So talk about it that way instead.
It sounds less revolutionary, but many of the most important social revolutions in history are really the aggregation of many small things rather than a big bang that blows everything apart. It’s a safer path to change as well, and it isn’t an act of surrender to say so. It’s an act of confidence and courage in where we are heading.
 Polanyi, K. (2001). The Great Transformation: The political and economic origins of our time. Boston, MA: Beacon Press. 6th edition
 Markoff, John (2022). Whole Earth: The Many Lives of Stewart Brand. NY: Penguin Press.
 Coase, R.H. ‘The Problem of Social Cost’, The Journal of Law and Economics 3. October 1960
 Levine, Rick et al. The Cluetrain Manifesto: The End of Business as Usual. New York: Basic Books, 2000.
 See for example report of 2021 polling by Public Affairs Council and Morning Consult, at https://newsdirect.com/news/national-poll-shows-public-trust-in-big-tech-is-falling-fast-161260638
 World Bank data on remittance transaction costs is at https://remittanceprices.worldbank.org/en
 This is often referred to as ‘The Minsky Paradox’
 Williamson, Oliver. (1975). ‘The Economics of Organization: The Transaction Cost Approach” American Journal of Sociology 87.
Cover Image: Veerle Contant on Unsplash