Governance Revolution
Dedicating my energies to the study of the social organization which is in the future to replace the present condition of things, I’ve come to the conviction that all makers of social systems from ancient times up to the present year, have been dreamers, tellers of fairy-tales, fools who contradicted themselves, who understood nothing of natural science and the strange animal called man. (Fyodor Dostoevsky, The Possessed)
Thomas Malthus is famous for missing the industrial revolution and predicting doom when plenty was around the corner. He was unlucky, publishing in 1798, when the revolution under way may not have been obvious. There’s some debate as to when things took off, some arguing it didn’t begin until the 1780s, others about 20 years before that. In any case, it seems clear that rapid growth, led by the textile industry, had to wait until the 1780s, while growth led by the steam engine occurred mostly after 1800. In general, the full effects of industrialization were only felt around the 1830s or so. Malthus just timed his book poorly.
I contend that we are in a similar position today with regards to a governance revolution under way. The Malthuses of today warn of insurmountable political deadlock, the dearth of community, and the crisis of democracy. Some routinely lament that America (and some other developed nations) is stuck in an institutional makeup that is hundreds of years old, hopelessly ill-adapted to contemporary challenges. Others insist that its very survival is proof of the immense wisdom of its framers, cautioning against any radical update. I think that both camps, in a way, lack imagination about the nature of institutional innovation.
In this post, I first present a few examples of what I mean by institutional innovation. Then I suggest that certain social environments enable greater innovation in governance, and illustrate with several historical examples. After that, I walk through a series of recent deep trends that I argue will make our social environment more prone to institutional experimentation, eventually leading to a governance revolution, with ramifications similar to those of the industrial revolution. Finally, I sketch some directions in which I anticipate considerable experimentation.
What is institutional innovation?
Before I dive into why I think a governance revolution is under way, I first want to give a sense that institutions have not, in fact, been that static over time. “Institutions” encompass more than just what kind of constitution you live under. While the writtenness of constitutions was itself a great innovation that spurred a process of continual improvement and increasing franchise, institutional innovation can take on many other forms.
For instance, for the longest time, elections in America almost always used the open ballot system, whereby voters voted by voice or by ticket color-coded by party, so that everyone could tell who everyone else was voting for. It wasn’t until the 1890s that secret balloting became common — an Australian invention imported to the US. That in my eyes constitutes institutional innovation.
Another example is public debt. The establishment of a systematic, large-scale public debt was a British invention of the 18th century. It went hand in hand with the founding of central banks. A national debt allowed a state to anticipate future tax revenues in situations of sudden increases in expenditures such as a war. And a central bank allowed the state to fund that debt on the cheap via privileged lending. Ferguson (2001) even argues that a national debt did a lot to broaden the franchise: giving taxpayers and bondholders representation actually made them more willing to finance public spending. I find the chart below quite striking; I wouldn’t be surprised if this British “invention” turned out to have played a major role in their taking over the world.
Private places are yet another example of institutional innovation that is not narrowly legalistic. Nineteenth century Great Britain experienced extreme demographic growth and urbanization rates: in 1800, the country had 11M people, a third of which lived in cities; by 1850, the population swelled to 21M, half of which lived in cities. This huge influx of people had to be housed with no apparatus of urban planning or land use laws and regulatory bodies, and no public supply of housing. The innovation was the creation of private submunicipal quasi-governmental organizations incentivized to provide public goods for the community, before selling individual lots with covenants (private constitutions) or leasing them. This was the origin of proprietary communities, known today as “common interest developments”, to which about a fifth of the American population belonged by 2010.
The invention of joint-stock companies, limited liability, trade unions, public funding of fundamental research, etc., all these are also institutional innovations. The point I’m trying to make here is that institutional innovation is broader than legislation and can be initiated by the public sector, the private sector, or a partnership between the two.
Williamson (1996) calls “legal centralism” the belief that governance necessarily comes in the form of centralized, publicly-run, legalistic mechanisms. Too often, people believe order couldn’t be supplied, or human behavior constrained, if it weren’t for the access to legal remedies (litigation, prison, etc.). In fact, history is replete with successful examples of private governance, as Stringham (2015) documents. In particular, he discusses the invention of financial derivatives — arguably incredibly important tools of economic governance — and shows that their enforcement was completely extralegal for a long time, proving that legal centralism is not a requirement, even in the case of contracts eminently prone to unilateral defection.
Friedman (2019) gives another fascinating example of bottom-up governance at work. In 19th century England, before the days of public police and public prosecution, crime victims were responsible for prosecuting their own case. While a successful prosecutor might get their expenses reimbursed by the court, it was unlikely to cover all expenses. A would-be felon, knowing this, might be tempted to commit their crime. The solution was for potential victims to deter crime in the first place by credibly committing to prosecute. This was achieved by joining an association for the prosecution of felons, contributing to a common pool, and then advertising one’s membership far and wide via newspapers. David Philips estimates there were somewhere between 1000 and 4000 such associations around 1839.
There are subtle differences in the few examples of institutional innovation I just presented. Some innovations are more localized than others. Democratic societies have many different elections at different levels of government, hence there is a lot of scope for trying out new methods, such as secret balloting. Similarly, there are always many proprietary communities simultaneously being developed, so that the fallout of bad ideas, for example in the design of covenants, remains limited. In contrast, the invention of central banking and the monopoly on paper currency, while definitely innovative, are an inherently totalizing experiment; and some might argue it did not benefit us overall.
This is not to imply an oversimplified picture of bottom-up good, top-down bad. For instance, mandatory vaccinations are an eminently top-down form of governance and yet usually considered a good thing. In the late 19th, early 20th century, there were still occasional epidemics of the terrible disease of smallpox. And mandatory vaccinations were no joke: “Public-health departments would send out teams of vaccinators — very often in the middle of the night — into tenement districts usually inhabited by, you know, immigrant, working-class people. They go door-to-door on these sort of vaccine raids, and they’d inspect the arms of everyone who lived in these homes to see that they had been recently vaccinated, that they had a kind of vaccine scar on their upper arms. In his [Lutheran minister Henning Jacobson who unsuccessfully sued the government to strike down mandatory vaccinations] own community of Cambridge [Massuchusetts], people are jumping out of windows and running the other way, or are getting doctors to sign phony vaccination certificates. I found one episode in the historical record from Kentucky, where the vaccinators went into an African American neighborhood of this community, and ordered everybody to get vaccinated, and those who refused were handcuffed and vaccinated at gunpoint.” Talk about extreme top-down government intrusion. And yet, arguably a good thing? Maybe?
In any case, top-down or bottom-up, public or private, localist or totalizing, we need a framework to think about and understand which social configurations favor successful experimentation in governance.
When to be a radical?
Brad Taylor builds upon Hirschman’s classic exit vs voice model to develop just such a framework, useful to determine what’s a safe environment for radical institutional innovation. His argument is pretty simple: the presence of “exit options” reduces the downside risks of reform by limiting them to the value of the next-best option. I want to emphasize here that I’m not just talking about exit as a tool to discipline one’s government (that effect may in fact be marginal). I’m talking about exit as a tool fostering both private or public institutional innovation, as indicated above. Every such innovation is a gamble, and every conservative since Edmund Burke has tended to warn against it.
You can actually formalize Burke’s stance by modeling the value of some institutional arrangement as a function of its quality. Assuming that the value function is concave leads a decision maker to act as if they were risk-averse and eschew reforms with similar probability of success or failure. Introducing an exit option, though, sets a floor on the value function, breaks the gamble’s symmetry, and changes the social calculus.
Of course, the above model is highly stylized and assumes that the people being experimented on can costlessly change jurisdictions in case of unfortunate outcomes.
Exit options are not just useful for experiment subjects. They are also useful for experimenters. Indeed, they allow social entrepreneurs to market their reforms to different jurisdictions. This is what enabled Columbus to propose his bold plans to different crowns of Europe. This phenomenon arguably played an important role in the Renaissance. Joel Mokyr talks about “cultural entrepreneurs”, the likes of Descartes, Grotius, or Erasmus, members of the cosmopolitan “Republic of Letters”, pushing some pretty radical agendas. One wonders to what extent the system of competing European states allowed them to survive by switching from one to the other. This “exitability” may have had a big impact on the persistence and development of a culture of contestability in Europe, championed by those cultural entrepreneurs.
The shopping mall model
Shopping malls make for a good model of radicalism-via-exitability. That model also emphasizes that institutional innovation is not just in matters of public governance, but also encompasses private governance.
A shopping mall can be seen as a mini-country, where tenants and patrons are the citizens and the mall owner is the sovereign. A mall has public goods such as bathrooms and security, that need providing by the sovereign, who in turn needs to raise funds from the citizens somehow. When shopping malls started becoming popular in the 1960s, one imagines several business models were rapidly tried. Exit in this case is very easy: tenants can close up shop since alternatives were legion; similarly, patrons can just stop showing up. This is a very safe environment for radical experimentation, which typically leads to rapid convergence.
And indeed, as an example of convergence, it pretty quickly became standard for mall owners to “tax” their “citizens” in the form of a land-value tax: tenants pay rent proportional to the surface area they occupy. Mall owners do not tax transactions or income: just land value. Incidentally, many economists seem to agree this is the best kind of tax. The fact that it’s scarcely used in regular jurisdictions may say much about the intensity of the forces at play in governance innovation amongst regular governments.
Classical Greece
Organizational change tends to happen at the population level, not the individual level. German physicist Max Planck famously quipped that science advances one funeral at a time. Similarly, exitability enables institutional innovation.
The case of Ancient Greece is edifying in that respect: as an ecology of about a thousand independent city-states, it was an environment ripe for rapid institutional innovation. From 800 to 300 BCE, per capita consumption doubled and the population grew by an order of magnitude (x10–15). According to Ober (2015), institutional innovation was at the core of this efflorescence. The spread of fair rules, in the form of formal institutions and cultural norms, promoted capital investment and lowered transaction costs, inducing people to invest in themselves in the form of literacy, numeracy, military training, etc. Competition between cities ensured successful institutional innovations “spread by learning and adaptive emulation”, facilitated by a shared culture and language. This led to innovations such as federalism, standard coinage, euergetism, some social insurance (grain price stabilization, welfare for invalids, upbringing of war orphans), epigraphy (etching laws on buildings), market regulations, adjudication systems, selection procedures to populate city councils, etc.
It’s worth noting that half the population growth in classical Greece occurred via colonization of often virgin territories, hinting at the importance of exit for radical innovation. The mobility of Greeks is also made manifest by the high share of foreigners in the population of the bigger cities (about a quarter of the non-slave population in Athens), and the fact that many rights were extended to them (eg, in many cities, noncitizens were allowed to litigate, own real estate, run non-governmental organizations, and enjoyed free access to some public services).
Renaissance Italy
The rise of medieval cities, especially in Italy, is another example of a similar ecology at work. While nominally controlled by a “holy Roman emperor”, many Italian cities managed to get charters granting them increasing levels of independence, paying lip service to the Emperor while de facto self-governing.
Thanks to their prosperous trade and their cleverly playing the Pope against the German king, the Italian towns managed to repel many invasion attempts. Dealing in high profit margin trade such as precious textiles and rare spices, they could stand on their own against opposing interests. Spruyt (1994) indicates that profit margins in the northern seas ranged around 5–15% (sometimes 25%), whereas those of the Venetians and Genoese could reach 150%. Northern Italy was also one of the most urbanized areas of Europe, with up to 30% of the population living in cities. Venice at the end of the 15th century had a population of almost 1.5M, comparable to the population of the Dutch Republic in the 16th century. Despite having a tenth of France’s population, the Venetian government had a bigger budget. Each of the Italian city-states managed considerable armies and navies. This prosperity, though, came with a “life-and-death struggle” between the city-states to “control ports of access in the Black Sea and the Mideast” (Spruyt, 1994).
Just like the poleis of classical Greece, the ecology of Renaissance Italy, and its density, favored mutual competition rather than centralization, leading to its own well-known economic and cultural efflorescence. And while the Italian cities did to some extent eventually consolidate into a few dominant actors such as Florence, Venice, Genoa, and Milan, many of the annexed towns maintained much of their previous independence, with considerable power devolved to local government such as statutes, councils and offices — not unlike the loose federalism of classical Greece with the domination of a few poleis like Sparta, Syracuse, and Athens. In both cases, this created an environment prone to institutional experimentation.
American Frontier
Likewise, moving-frontier America offered many opportunities for experimentation in governance. This paper speaks of a “competitive struggle for citizens” with highly mobile settlers sensitive to governance alternatives. Even in 19th century America, with its great expanse of arable land and its chronic shortage of labor, the expanding frontier continued to play a role in institutional innovation, especially in matters of civil society. As Wiebe (1995) puts it, “the frontier freed Americans to create their own society.” They developed a “rugged, adaptable, home-grown” democracy that “crowned the achievements of the Western World… In fact, the frontier itself only made sense as development, as the ceaselessly repeating process through which democracy renewed itself.”
Tocqueville reported extensively on this incessant civic hustle and bustle:
Americans of all ages, all stations in life, and all types of disposition are forever forming associations. There are not only commercial and industrial associations in which all take part, but others of a thousand different types — religious, moral, serious, futile, very general and very limited, immensely large and very minute. Americans combine to give fetes, found seminaries, build churches, distribute books, and send missionaries to the antipodes. Hospitals, prisons, and schools take shape in that way. Finally, if they want to proclaim a truth or propagate some feeling by the encouragement of a great example, they form an association.
In the few examples I presented, we have an intensely competitive environment and relatively abundant exit options for the polity. The availability of institutional alternatives both served to empower the people and to foster emulation.
Exit in a closed world
The frontier, though, has since closed. And most of the world population has been consolidated into large nation states rather than ecologies of relatively small political units. So what does exit look like in a “closed” world? Is this still an environment amenable to radical institutional innovation?
My thesis is that exit is, in fact, becoming easier, and has taken a form much different from that of a moving frontier. This is not to say that exit should be desired for its own sake. Glen Weyl convincingly warned us of the dangers of fetishizing exit, derisively calling its biggest proponents “exitocrats”. But exit should be embraced to the extent that it enables experimentation.
There are four main trends that make exit easier and therefore enable greater institutional innovation.
1. A greater share of wealth in mobile assets. The share of US household wealth held in financial assets has grown significantly over the past decades. For instance, the share of corporate equities in household portfolios has doubled since 1950, from about 20% to about 40%. McKinsey reports that, in 2020, financial assets held by households, governments, and nonfinancial corporations around the world, were similar in size to real assets (about $500T). Wealth in cryptoassets peaked at $3T in 2021. While amounting to less than 1% of total world wealth, it is a new and growing asset class. Financial and crypto assets being inherently more mobile than real assets, their growing share of total wealth should translate into increased general mobility. Crypto wealth, when properly managed, holds the further promise of being really hard to confiscate, making it even more mobile than traditional financial wealth. Its inflation-resistant properties allow one to opt out of the need to invest one’s income into risky asset classes such as corporate bonds and equities. Indeed, inflation corrals wealth into the surveillance pens of traditional financial assets which typically require third-party custody and as such are more easily confiscated. Bitcoiners like to dub this a “6102 attack” in reference to Roosevelt’s 1933 executive order that effectively seized all gold held by the American public — made possible by the fact that gold was likely to be custodied by a third-party such as a bank. Generalizing this idea, any asset class that is less likely to be self-custodied, such as stocks, bonds, or metals, is de facto less mobile because it is at greater risk of seizure. In contrast, the asymmetric cryptography at the heart of cryptoassets renders large-scale seizure cost-prohibitive. By freeing their holders from the yield-chasing imperative, cryptoassets turbocharge wealth mobility.
2. Deterritorialization of goods and services. Many goods and services have become location-agnostic. Supply chains have become a sort of abstract global infrastructure, with Amazon Prime bringing any good from anywhere, even food, to your doorstep. The internet has turned entertainment into a largely virtual phenomenon. The pandemic gave a boost to both telehealth and online arbitration services. New models of schooling such as Synthesis seem to be fundamentally indifferent as to the physical location of their students. Although traditional finance still largely dominates the sector, there is a growing sense of the long-term inevitability of its deterritorialization by migrating much of its functionality to blockchains, making bank branches a thing of the past. Even online-first socialization seems to have stepped up recently: while forums and chat rooms have a long history already, online social experiences have grown more multidimensional recently, such as interintellect.com, a platform that runs live, video-based online salons, and maintains a well-organized Discord server that feels like a virtual city. In the realm of public goods, Estonia has put 99% of its administrative services online (from voting, to filing taxes, authenticating, registering newborns, digital signatures, etc.). With its e-residency program, Estonia advertises itself as a “borderless digital nation for global citizens” and talks about “transnational digital identity”. Virtual governance is nothing new. Churches, mutual aid societies, merchant networks, artisan guilds, employers, professional associations, healthcare organizations, cryptocurrency networks — all to some extent provide club goods to their members at a scale that can transcend or cut through territorial jurisdictions. Even formal legal systems have existed, and still exist, independently of territorial systems of law, such as the Law Merchant in medieval Europe (to which one can trace the lineage of contemporary international law). If the production of goods and services becomes more location-agnostic, and if infrastructure becomes inherently more abstract and global (supply chains governed by international law, internet, blockchain, etc.), mobility is enhanced.
3. Cities as the preeminent political units. Long-term trends point toward ever-increasing urbanization rates, with a disproportionate share of GDP originating in cities, despite pandemic-fuelled prognostications about the death of cities. Traditionally, the phenomenon of cities is described as an efficient way for job seekers to find job offers. Cities have been seen primarily as job markets. But some commentators are arguing that a city’s amenities, not its jobs, are becoming the primary reason for moving there. Just like the democratization of the automobile turned hotels from being mostly about the right location with lots of foot traffic to being mostly about branding, the deterritorialization of economic production may be pushing cities to evolve from being job markets to being branded consumption bundles. As commercial leases vacate, that space opens up for both lifestyle-related activities and last-mile infrastructure for the abstract global supply chain. You may no longer be looking to live in the city primarily for earning a livelihood but for the lifestyle it affords. Economic agglomeration is replaced by affinity-based agglomeration. It’s no wonder that some city mayors seem to have taken to some kind of branding exercise. This phenomenon, in turn, may lead to greater population sorting, greater heterogeneity in the vibes of different cities. There is a sense in which cities are becoming the preeminent political units and the cornerstones of one’s political identity. Cities around the world have even started working together to address common problems independently of their host nations. And when the place one lives in becomes a lifestyle statement, mobility becomes more salient.
4. Portable identity. Identity is as central to modern government as territorial maps. In particular, as Scott (1998) argues, a fixed, permanent identifier is necessary for things like “tracking property and inheritance, collecting taxes, maintaining court records, performing police work, conscripting soldiers, and controlling epidemics” — all things requiring correlating identity across contexts, for which a fixed, central identifier is the natural solution. And yet, widespread permanent patronyms are a relatively recent phenomenon, and originate in state projects that, throughout history, have encountered popular resistance, mostly because the initial impetus was primarily fiscal. Traditionally, identity was quite flexible: individuals could have different names at different stages of their life, in different contexts, or with different interlocutors. In short, identity was highly localized and contextual. With the growth of state-imposed patronyms or surnames, Scott (1998) hypothesizes that people for a while kept a double identity: a fixed (often totally made-up) official one and a potentially changing local one — the former eventually winning out via a combination of coercion and incentives (avoid double taxation and/or claim an inheritance). Just like taxation, identification is only voluntarily embraced if it involves a beneficial trade, namely political representation, or the provision of genuine public goods, such as enhancing public health (tracking epidemics often requires tracking people), busting local monopolies of information, or lowering transaction costs. Modern identification, however, crucially depends on a central provider, limiting mobility beyond its purview. In contrast, recent innovations like public key cryptography and blockchain technology allow anyone to self-assert their identity and make provable claims about themselves without the need for a central identity provider. And while names may feel special, the reality is that in many cases they originated in bureaucratic fantasy and are no less arbitrary than the gibberish of a cryptographic public key. The main difference is that the latter is portable across administrative systems while the former is akin to being stuck in a proprietary database.
The future of governance?
The above four trends point to greater exitability in the future, hence, I posit, greater experimentation in governance. Detailed explorations of those possible experiments will be the subject of future posts. In this section, I will merely touch upon a few broad avenues of experimentation that I foresee.
Layered governance
I mentioned earlier the growth of common-interest developments (CIDs). A related structure is that of special economic zones (SEZs). The main difference between the two is that the former constitutes a soft fork of the background governance (it’s only adding constraints or rules) while the latter constitutes a hard fork (it is removing constraints). SEZs are specially carved zones where not all normal rules apply. Historically, ports were the most common such zones, and the hard fork almost always consisted in some kind of freedom from duty (tax exemptions), from the free port of Delios, established by the Romans in 166 BCE, to colonial outposts such as Macau or Hong Kong, to contemporary export processing zones. SEZs can take on a broader meaning though. For instance, entrepreneurial settlements in the New World under royal charters in the 17th century can be seen as proto-SEZs.
Properly speaking, the first modern SEZ was established in 1959 in Shannon, Ireland. Since then, the concept has experienced a dramatic growth, especially in developing countries after 1985. By 2018, the world as a whole had around 10k SEZs, and 75% of all countries have at least one SEZ. They are also growing in sophistication. From offering merely exemptions from some customs duties, they are becoming more encompassing, with special regimes for industry, commerce, residential sector, tourism, etc. They are becoming more like real cities in terms of scope, as well as more privately-owned, developed, and operated. That said, SEZs are no panacea, and some argue they have mixed results overall, especially in Africa. Nonetheless, I think the growth of both CIDs and SEZs constitutes an interesting trend.
Combining the growth of SEZs with the increased political prominence of cities, I foresee some experimentation in layered governance. Much like blockchains are layered, with a “layer 1” used for settlement, and a “layer 2” used for amortizing multiple operations, could we have a “layer 1” of legacy governance, territorially defined, and a “layer 2” of governance that is more location-agnostic?
With enough bargaining power, extant cities could obtain concessions from host nations and hard fork in a way similar to SEZs. And existing city alliances could be the starting point for the harmonization of a new, nonterritorial layer of governance linking cities together in a kind of virtual polity, with SEZ charters as the anchor points into legacy governance systems. When much of government is “software”, its virtualization in layer 2 is made possible, as the example of e-Estonia shows. And if big companies, which are a mix of software and hardware, can go fully remote, why couldn’t city leagues? Estonia has backed up its government servers in Luxembourg (and plans to have 4–5 more such “data embassies” around the world) so that, in case of invasion and loss of territory, it can boot itself back up as a purely virtual country. Layer 2 governance, while firmly anchored into layer 1 governance, could nevertheless decouple to some extent, and we could witness interesting reconfigurations at this new layer. We could even witness some form of unbundling of government, where the modules that make sense to tie together are tied together into a “jurisdiction” and belonging to multiple jurisdictions becomes commonplace. Portable identity would further enable such fluidity.
Multinational companies come close to what I described above. After all, Google has offices all over the planet, and a Google employee can fly into any of those, badge in with their “universal” Google badge, and enter a a world of familiar infrastructure and community. But commercial companies are not civic organizations. A better historical equivalent are mutual aid societies, which flourished in the UK and the US in the 19th and early 20th centuries. In the UK, they were mostly known as “friendly societies”, while in the US they were more often referred to as “fraternal orders”. In both cases, they were voluntary not-for-profit societies for reciprocal relief (sick pay, life insurance, healthcare insurance). In a nutshell, they were what welfare consisted in before the advent of the welfare state.
More often than not, those societies took the form of a federation of local self-organized branches that spanned a whole country and in some cases were transnational (across the pond). Their strong emphasis on character-building, regalia, founding mythology, and ritualism, formed a culture more akin to a civic identity than to the mercenary identity of enterprise employees. Mutual aid societies enjoyed widespread popular membership across all socioeconomic classes. David Green (1993) argues that, of the 12 million people covered by Britain’s National Insurance Act of 1911 (which nationalized social insurance), 9 million were already covered by mutual aid schemes. David Beito (2000), speaking of the experience of fraternal orders in the US, estimates that “one of three adult males was a member in 1920, including a large segment of the working class.” He further argues that fraternal orders were far from a native white phenomenon. Au contraire, they “achieved a formidable presence among blacks and immigrant groups from Eastern and Southern Europe.”
Mutual aid societies exhibited great variety in governance structure. Some focused on sick pay, others on funeral benefits. Some provided healthcare via a fee-for-service scheme, others structured it in a capitated format. Some set up district-level medical institutes with a choice of doctors, others opted for local branches contracting a single physician. Some graduated their contributions by age at initiation, some had a flat rate. Some were more federalized than others. All together, they formed a landscape of competitive governance that exhibited tremendous capacity for innovation and adaptation. Their rulebook was deep and the importance of the societies in the members’ everyday life is undeniable. As such, they truly constituted a second layer of governance, a soft fork of the limited government background of 19th-century Britain and America.
A similar logic of mutual aid and self-help is behind another major historical example of a nonterritorial second layer of governance: the Hansa. The Hanseatic League was a formal league of cities in Northern and Central Europe that existed roughly from the 14th to the 17th century. At first, it was just a group of merchants acting together to protect their commercial convoys against pirates, independent of territory. Over time, as towns throughout Europe, and Germany in particular, gained independence and privileges, those local organizations of merchants started to develop similar institutions and became more linked to city councils. This matured into alliances between cities, for mutual assistance against the encroachment of German princes. At its peak, the Hansa encompassed 200 towns and was capable of signing treaties with sovereigns and even waging war at times.
Beyond defense agreements, the Hansa sought to build rational institutions for weights and measures, coinage, as well as toll systems. The league was useful for obtaining international trade privileges such as tax exemptions or monopolies on trade routes (forbidding access to ports for non-members), for instance with England or Flanders. The northern trade, in contrast to the Mediterranean trade of the Italians, was bulk (commodity) trade with low profit margins. This is why the League’s main focus was on reducing transaction and information costs and seeking monopoly power via exclusive trade deals with various sovereigns — in other words local “hard forks” in governance.
In essence, the League took on a role similar to that of the sovereign state in France at the time: that of a business-friendly rationalizing and standardizing force. Yet it was very different in form as it had no clear hierarchical authority or territorial borders. Member towns had territorial overlords (L1 governance), but they also enjoyed great latitude for self-organization (L2 governance). Some were fully independent, others were nominally under lordship but factually independent, others yet were still subject to crosscutting jurisdictions. The League developed its own institutions such as the Diet (or “tag”), a general assembly that convened usually every several years, “vot[ing] on membership, a wide range of economic matters, external representation of the Hansa, and military issues.” The Hansa could also “raise an army, conduct foreign policy, decree laws, engage in social regulation, and collect revenue” (Spruyt, 1994). The Hansa aggregated multiple regional groups of towns that shared similar interests, and each of these groups had their own treasury. These regional groups were sometimes further subdivided into smaller coalitions of towns.
Mutual aid societies and city leagues are far from perfect illustrations of what I mean by layered governance. Mutual aid societies illustrate the benefits of competitive governance in layer 2. City leagues illustrate nonterritorial sovereignty. Both show the production of voluntarist state capacity. Experiments in the virtualization of governance in the future are likely to look very different. But these historical examples serve to show that territorial governance isn’t necessarily “natural” or inevitable. Alternatives have existed and could return.
Proprietary governance
Another direction in which the Governance Revolution could go into is the rise of proprietary governance. Rather than layering up governance, we could witness its increased privatization. The key underlying assumption here is that the optimal size of the sharing group for many public goods is actually smaller than many people believe, even smaller than the typical municipality. This in fact — so the argument goes — is why CIDs, which are submunicipal voluntary associations, naturally emerge. But that is not the only option.
There are essentially two broad types of proprietary governance: (1) fractionated ownership, and (2) unitary ownership. The first one is what most people are familiar with, in the form of homeowner associations, condominium associations, or cooperative residential communities — all variants on a logic of collective private ownership. They are effectively contractual governments that provide public goods (roads, sewers, parks, police protection, etc.), “tax” residents (via “assessments”), and elect their officers. Typically, such communities are initially built from scratch by a single subdivider that sets up a “constitution” (covenants or, in other words, private zoning), then sells the individual lots. The single subdivider is incentivized to design the constitution well so as to maximize the value of the parcels to be auctioned off. The constitution of such “contractual governments” typically differs from that of traditional governments, first and foremost in the fact that voting is rarely one-person-one-vote and more likely to be one-property-unit-one-vote or proportional to unit value. Much of Great Britain’s urbanization in the 19th century followed that playbook of a single subdivider and private zoning.
Spencer MacCallum, in Beito et al, (2001), however, criticizes fractionated ownership, and describes these “neighborhood governments” as “often arbitrary, unresponsive, and dictatorial”, encumbering development with much greater restrictions than municipal governments. He points to regulatory distortions for their success, not consumer demand. Indeed, federal tax policy penalizes leasing for residential use and there are various federal subsidies for subdivision housing. Instead, he proposes to substitute the logic of collective action with that of the residual claimant: a single entrepreneurial landlord develops the land and leases various units. Good governance and good infrastructure increase the value of the land, which in turn increases the value of the proprietor’s residual claim. The success of the land lease model is evident in commercial real estate such as office parks, medical clinics, shopping centers, marinas, research parks, etc. Businesses and professionals rarely own their premises, because land owning and land management require specialized administrative services. Such proprietary administration, MacCallum argues, would emerge just as naturally for residential real estate, if it wasn’t for the fiscal distortions that end up favoring collective private ownership schemes. Adam Smith, in The Wealth of Nations, seems to concur:
The interest [of landowners] is strictly and inseparably connected with the general interests of the society. Whatever either promotes or obstructs the one, necessarily promotes or obstructs the other. When the public deliberates concerning any regulation of commerce or police, the proprietors of land never can mislead it, with a view to promote the interest of their own particular order; at least, if they have any tolerable knowledge of that interest.
In other words, unitary ownership aligns the incentives of the proprietor with those of the residents particularly well. It is the shopping mall model writ large. MacCallum describes this as the transformation of government “from a political process to one purely economic”. Land becomes a capital enterprise. In contrast, MacCallum argues that land in CIDs is inherently more speculative inasmuch as the value of the lots depends on factors mostly outside the direct control of each individual owner, due to the collective nature of decision-making. If land value increases, more often than not it’s due to artificial scarcity, rather than true improvements by the community. He further argues that incentives are less aligned between a subdivider and the individual buyers, as the latter are typically less sophisticated and therefore more prone to be taken advantage of. This, however, is not a concern in a land lease setup.
The biggest problem with MacCallum’s vision of proprietary governance is that, while a resounding success with commercial real estate, it scarcely has been tested with large-scale residential real estate (apartment complexes don’t really count here). Company towns, in particular mining towns, are probably the closest historical precedent, but they are far from a perfect analogy, as their isolated nature and their single employer made them quite unlike the typical city.
One may argue that colonialism approaches MacCallum’s vision, but there are important caveats there. Most scholars agree that colonialism was mostly bad. That said, Acemoglu and Robinson’s reversal of fortunes thesis is a useful framework to distinguish between different types of colonialism. According to this thesis, countries that were relatively poor when they were colonized tended to perform better after the colonizers’ departure, and vice versa for countries more prosperous to begin with. And less developed areas fit more closely the situation of an entrepreneurial landlord looking to develop a community from scratch. Similarly, the size of the colony is another important factor if we are to draw an analogy. The larger the colony, the more captive its population, and the greater its weight in global capital markets. Instead of flying in case of trouble, capital would be more incentivized to employ coercion of some sort to quell it.
All in all, Hong Kong, as a small colony that started off with essentially zero natural resources or pre-existing economy, is a potentially valid illustration of MacCallum’s vision, with Great Britain as the “residual claimant”. Needless to say, Hong Kong experienced an economic miracle under British rule, going from a barren island in 1945, with an income per capita 30% that of Britain, to a thriving global metropolis in 2017, with an income per capita 40% higher than its former colonial overlord (Monnery, 2017). Hong Kong experienced chronic budget surpluses throughout its colonial rule, while tax rates were maintained extremely low all along. Generally speaking, the colonial administration seemed to have a genuine interest in the welfare of the residents. Ming K. Chan, writing in a 1997 review of the British legacy, observed: “Hong Kong today is globally recognized as a remarkable example of a liberal society with a vibrant economy where its population of more than six million enjoy their freedom and opportunity. For this, the British can indeed claim considerable credit.”
One of the forces that Lutter (2016) identifies as key to keeping a city proprietor “honest” is “the need to have good relations with the host country and citizens not residing in the city.” He adds: “There is always the possibility that the host country will renege on the initial agreement by forcibly taking control of the proprietary city.” In the case of Hong Kong, this can refer to the need to keep the mothership back in Westminster happy, lest the colonial administrators be fired. Alternatively, this can be interpreted as the looming threat of a takeover by mainland China, in which case weakening Hong Kong through plunder would have been running the risk of inviting an invasion. Yet another interpretation is that the administration was disciplined by the ease with which capital could flee Hong Kong in case of serious mismanagement, given its relatively small footprint in global capital markets — a reality that generations of financial secretaries emphasized time and again.
A final feature of colonial Hong Kong that further strengthens the analogy to a proprietary city is the fact that “land premium” made up a significant share of government revenue. This is in line with what theory would predict to be the main revenue source of a proprietary city: land value taxes and Pigouvian taxes.
Is colonial Hong Kong a good large-scale model of proprietary governance? Can its success be generalized to this type of governance? Hard to say, this is mostly suggestive. Generally speaking, the relatively limited geographic scope of proprietary governments further facilitates exit and fits in with the trend toward greater consumerism, rather than professional imperative, in choosing where to live.
The idea of cities as consumption bundles also ties back to the literature on the so-called Tiebout model, according to which people choose where to live based on the bundle of public goods offered in that location. Some places will have low taxes and low public goods, and other places will have high taxes and high public goods — and people vote with their feet. That is the essence of what MacCallum meant by turning government “from a political process to one purely economic”. There is an enormous literature on the Tiebout model, but generally speaking, we can expect that greater mobility and greater choice in jurisdictions will enhance the applicability of the Tiebout model.
In terms of mobility, I’ve argued that it is likely to keep increasing. In addition, a key assumption of the Tiebout model is that access to jobs is not a consideration for choosing where to live — something which I’ve argued is a deep ongoing trend, with cities evolving from job markets to consumption bundles. Choice in jurisdictions is likely to increase as the cost of setting up new jurisdictions is reduced by both technological developments and charter city-friendly legislative changes around the world. The greater the number of jurisdictions, the more difficult for them to collude (via, say, tax harmonization), and the likelier they are to compete for mobile “customers”. Will this market process for public goods benefit people? This is an immense debate. Dowding et al, in a 1994 review of the literature, conclude that “there is mixed but marginal support for the proposition that the greater the number of jurisdictions the greater the satisfaction level for some, though not all, locally provided collective goods.”
Taking the private governance logic to its extreme, we end up with anarcho-capitalism. This happens when private actors not only produce public goods such as infrastructure and amenities but also start “selling law” as just another public good. The distinction between public and private goods in effect falls apart here, as they effectively become excludable. Pure public goods are actually quite rare. Most public goods are in reality club goods, and a private entrepreneur can effectively supply them. Law could end up bundled with infrastructure, or it could be provided by specialized agencies that one would hire independently of infrastructure. In practice, those agencies are likely to be protection agencies that declare the rules they will enforce. This is not to say that we would end up in a patchwork of legal systems where two neighbors live under incompatible rulesets. There would be powerful economic forces toward legal homogenization, as there are economies of scale in law enforcement when a protection agency can increase its customer base. Homogenization will make protection cheaper, and customers are likely to compromise on their values for cheaper protection.
Anarcho-capitalism has a bad rap. But people too often dismiss it on superficial grounds, equating it with chaos and lawlessness, when its proponents in fact insist that it is highly orderly. According to them, rules and laws would still be very much present, and in fact day-to-day life wouldn’t look very different from daily life in existing nation states.
State-reinforced self-governance
Another general direction in which I anticipate lots of experimentation is in the return of voluntarism — but a voluntarism empowered by what Lin Ostrom called generalized institutional facilities. I have found various expressions out there to capture this trend, although the people who coined those terms may not agree with my interpretation of them. The one I use to title this section is from a paper describing community-based irrigation systems in Japan. Another one is state-capacity libertarianism, which Tyler Cowen proposes to transcend the fundamental paradox of libertarianism. “Civic capitalism” is the term used by David Green (1993). A final one is positive non-interventionism, coined by financial secretary Haddon-Cave to describe the governance style of the colonial Hong Kong administration, especially under the guidance of his predecessor, John Cowperthwaite. But perhaps the clearest characterization of this type of governance is in the collective works of Lin Ostrom, which she summarized as polycentric governance.
All those terms get at the same underlying reality. A story told in Lin Ostrom (1990) captures this reality well. It’s the story of a groundwater basin in 1960s California from which several companies were drawing water — a classic common pool resource problem. The companies would have faced the adverse consequences of basin depletion, had they not managed to cooperate and create innovative institutions to manage the basin. Ostrom celebrates the voluntarism of the arrangement, but the thing that struck me the most was the role of the government facilities in the whole process. The US Geological Survey and the California Department of Natural Resources produced knowledge about the geological reality of the basin that was crucial to the institutional innovations the groundwater pumpers devised. The pumpers did not have to cover the full cost of those technical studies. They also benefited from equity court proceedings, with part of the cost again covered by the state of California. Finally, some legislation was passed by the state legislature to organize several special-purpose districts and interdistrict arrangements.
Strong yet neutral, effective yet unopinionated, those facilities embody an ideal style of governance. Ostrom describes them as generalized institutional facilities, a term which I think perfectly captures the ethos of state-reinforced self-governance: problem-solving coming from the bottom and facilitated by the top — not driven by the top. Facilitation, not intervention. It’s worth emphasizing the importance of setting expectations for such a governance style. In a system where everyone expects the “top” to drive solutions to collective action problems, the groundwater pumpers might have been tempted to await government representatives on their fact-finding missions. Worse, they may have been incentivized to distort information to benefit themselves, “present[ing] the ‘facts’ of the local situation in such a way that officials who may not know the local circumstances well will be led to create institutions that will leave some individuals better off than others”, as Ostrom explains. “Those individuals who have the resources to enable them to make the best case to external officials are most likely to gain rules (or exceptions to rules) that will advantage them the most.” When, instead, the government simply provides mostly passive facilities, this fosters self-help rather than strategic behavior. Importantly, the nature of those facilities needs to be such that they are not cooptable. What I call government style, David Green (1993) calls government character, which can be captured by the question: “Are people being treated as means to the achievement of the government’s ends, or is the government making means available to people to pursue their own ends?” This, again, hints at Ostrom’s institutional facilities.
I also think this general philosophy maps to what Glen Weyl calls pluralism, which in some way is his own attempt at overcoming the chasm between progressives and libertarians. When I hear Weyl talk about institutions that foster “a wide range of latitude for broader social input”, I think about Ostrom’s generalized institutional facilities. In the case of Weyl, he emphasizes participative democracy, urging us to evolve our “formalisms” from thin (votes, likes, webpage views, market demands) to thick (conversations, papers, seminar). He often points to Taiwan’s digital democracy as a relevant form of ongoing experimentation, with digital minister Audrey Tang at its helm.
Taiwan’s digital democracy movement had its start when g0v (“gov zero”) burst onto Taiwan’s tech scene during the 2014 Sunflower protests against a free-trade deal with mainland China. G0v was essentially the movement’s tech support, but it has since grown into an influential tech organization, all the while resisting institutionalization, insisting on remaining a “polycentric community of self-organized contributors”, as per their manifesto. Mei-chun Lee calls it “parasitic in-betweenness”. Others might just call it a new kind of civil society, with a hacker twist. Their first impactful “hacking” project was the decentralized manual digitization of campaign finance records. Here I would liken the campaign finance records themselves to a government “facility” that empowered a voluntarist organization to build a useful product around it. With their Cofacts project they took aim at public misinformation by building a chatbot on a popular messaging platform that will give nuanced clarification about rumors. The chatbot’s responses are crowdsourced by volunteer editors. In this case, I’d argue g0v itself created an institutional facility. Next, they created AirMap, a real-time map of particulate matter, leveraging thousands low-cost sensors deployed by thousands of citizens, and open-sourcing the data. The tool, this report argues, “forms a far richer picture of air quality across Taiwan than any of the centralized government sensors or private sensors that had previously been installed in expert-selected locations.” The report further emphasizes that the project is a “collaboration between LASS (an open hardware community), Academia Sinica (a national research institute), private manufacturers (who later produce and sell their versions of AirBox using the coalitions’ prototypes), g0v (Taiwan’s civic hacker community), and the public sector, especially public schools”, again illustrating the virtues of voluntarism within the state’s generalized institutional facilities.
G0v cranked up their ambition with vTaiwan, a national-scale deliberation platform bundling collaborative technologies, survey tools, and facilitation techniques. The goal is to intelligently surface popular sentiment on various topics to eventually drive legislation and/or regulation, as they did in the case of the regulation of rideshare apps. While the platform may have lost some traction, its influence is felt in the ongoing Participatory Officers network, a more institutionalized version of vTaiwan, integrated into the policymaking flow, and led by digital minister Audrey Tang.
It’s no surprise that, to describe their movement, some of the main actors at the heart of Taiwan’s digital democracy claim the same term — “polycentric” — that Lin Ostrom used to describe her own work. In the conclusion, I explain why.
Conclusion
I’m hopeful greater exitability will foster greater institutional experimentation. Theory suggests the two are linked, and history seems to confirm that. Alongside technological changes, recent deep trends in the mobility of wealth, the production of goods and services, the political capital of cities, and the nature of identity, lead me to posit an imminent governance revolution. This revolution, far from a rapid affair, is more akin to the industrial revolution, which is arguably still ongoing, 200+ years in. In many ways, our social institutions have not been revolutionized anywhere near the extent that our scientific and engineering capabilities have. In fact, many consider this stability a good thing — an idea which this essay questions.
I suggested several forms that institutional experimentation may take, from layered governance creating a new, virtual, second layer of governance, to proprietary governance embracing entrepreneurial community building, to state-reinforced self-governance, emphasizing a style of government best described by Lin Ostrom’s notion of generalized institutional facilities. While the first two are pretty directly rooted in some of the deep trends I identify, the third one deserves perhaps a more explicit linkage.
In many ways, state-reinforced self-governance involves the blossoming of voice. How can that be, in a world of greater exitability? The thing is, exit and voice need not be substitutes. In fact, they often are complements. The problem of voice is that any individual voice has a vanishing probability of making a difference, whereas exit decisions have a guaranteed impact on the decision-maker, which naturally prompts them to get informed. Exit decisions are akin to consumption decisions: when you spend money on a good, you mostly know what you’re going to get, whereas when you spend votes on an election or a ballot, what you get depends on what all the other voters do. As a result, in general it pays to inform yourself before spending money but it doesn’t pay to inform yourself before spending votes. In other words, informed voting has tremendous externalities. David Friedman puts it very well in The Machinery of Freedom:
Externalities play an enormously greater role in institutions controlled by voting. If I invest time and energy in discovering which candidate will make the best President, the benefit of that investment, if any, is spread evenly among 200 million people. That is an externality of 99.9999995 percent. Unless it is obvious how I should vote, it is not worth the time and trouble to vote “intelligently”, except on issues where I get a disproportionately large fraction of the benefit. Situations, in other words, where I am part of a special interest.
This is why collective action (voting) often ends up being more guided by evolved beliefs or expressive preferences, rather than rational evaluation. Schumpeter, in Capitalism, Socialism and Democracy, paints that dreadful portrait:
Thus the typical citizen drops down to a lower level of mental performance as soon as he enters the political field. He argues and analyzes in a way which he would readily recognize as infantile within the sphere of his real interests. He becomes a primitive again. His thinking becomes associative and affective.
But if you can get the same outcome via either exit or voice, the information you gathered for the former can carry over into the latter. “When there is overlap between the informational requirements of exit and voice decisions,” Brad Taylor argues, “exit will increase the epistemic quality of voice.” If nothing else, it gives voice alternatives to point to, relieving from the costly need to theorize those alternatives. More generally, the option to exit something sharpens the mind because it becomes worthwhile to understand the thing. As an example, and speaking personally, the advent of Bitcoin has definitely sharpened my thinking about the monetary system and increased the epistemic quality of my voice on the matter, precisely because Bitcoin has given me the option to exit the current monetary system. And this is true whenever individual exit and collective voice co-exist, from patient choice in healthcare, to quasi-markets in pensions, private alternatives to public broadcasting, and many more. This is why I’m hopeful that greater exitability will generally enhance voice, diminishing the need for technocracy, and why I believe generalized institutional facilities can turbocharge voice. Dostoevsky may be right to call would-be social reformers dreamers and fools, but staying quiet is not a viable alternative.
References
I only include the references that I couldn’t hyperlink to in the text.
Beito, David T., et al., The Voluntary City, The Independent Institute, 2001
Beito, David T., From Mutual Aid to the Welfare State: Fraternal Societies and Social Services (1890–1967), The University of North Carolina Press, 2000
Bell, Tom W., Your Next Government? From the Nation State to Stateless Nations, Cambridge University Press, 2016
Ferguson, Niall, The Cash Nexus: Money and Power in the Modern World, 1700–2000, Basic Books, 2001
Friedman, David, et al., Legal Systems Very Different From Ours, 2019
Green, David, Reinventing Civil Society: The Rediscovery of Welfare Without Politics, Civitas, 1993
Lutter, Mark, Three Essays on Proprietary Cities, George Mason University, 2016
MacCallum, Spencer Heath, The Enterprise of Community: Social and Environmental Implications of Administering Land as Productive Capital, The Journal of Libertarian Studies, 2003
Monnery, Neil, Architect of Prosperity: Sir John Cowperthwaite and the Making of Hong Kong, 2017
Ober, Josiah, The Rise and Fall of Classical Greece, Princeton University Press, 2015
Ostrom, Elinor, Governing the Commons, Cambridge University Press, 1990
Scott, James, Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed, Yale University Press, 1998
Spruyt, Hendrik, The Sovereign State and Its Competitors, Princeton University Press, 1994
Stringham, Edward Peter, Private Governance: Creating Order in Economic and Social Life, Oxford University Press, 2015
Wiebe, Robert H., Self-Rule: A Cultural History of American Democracy, The University of Chicago Press, 1995
Williamson, Oliver E., The Mechanisms of Governance, Oxford University Press, 1996
