In 2009, economist Paul Romer articulated a vision for chartered cities -- new urban zones built on uninhabited land with fresh governance rules, free from the institutional inertia of existing nations. NullCase saw in this vision a striking parallel to what he was already doing in Minecraft, and he spent years developing the argument that sandbox games were not merely analogies for governance experiments but genuine testing grounds for them.

The Paul Romer Connection

Romer identified three requirements for a chartered city to flourish: good rules specified in a charter, uninhabited land, and partnerships between nations. NullCase mapped these directly onto the Minecraft server model. A server operator specifies rules (the charter). A newly generated Minecraft world provides uninhabited land. And the voluntary participation of players from around the world constitutes the partnership.

Liberty Minecraft embodied this framework. NullCase described the setup in concrete terms:

Liberty Minecraft is the size of Manhattan. Our world has just one rule: resolve nonviolent disputes nonviolently. There are two terms of use: do not hack the server, and read and understand the rules. Private property is protected by computer code. When non-subscribers leave for more than 60 days their property is auctioned.

Within this charter, the server produced functioning markets, private commons, transit infrastructure, monetary systems, and spontaneous community governance -- all the features that Romer hoped chartered cities might develop, at negligible cost and zero risk to human life.

Why Games Work

NullCase made a forceful case that games -- Minecraft in particular -- are vastly underestimated as tools for social and economic research. His argument rested on several practical advantages:

Games are more capable than you can imagine. Today players are building Turing complete computers inside Minecraft worlds with iPhone interfaces that display video calls. There are 10¹⁸ possible Minecraft worlds.

The economics were compelling. A high-quality Minecraft server cost less than $1,000 per year. The barrier to entry was so low that five-year-olds had operated their own servers. The risk was minimal because no one's livelihood was at stake. And the observability was total -- every transaction, every construction, every interaction could be logged and analyzed.

Perhaps most importantly, rules could be split-tested. NullCase demonstrated this with the private property experiment, where two otherwise identical areas of the map received a single rule change -- one had private property protections, the other did not -- and the outcomes were compared over months and years.

Rules can be split tested by using two otherwise identical worlds. It's about smart capital management.

What Emerged in Practice

The results from Liberty Minecraft were not abstract. Real phenomena emerged that had theoretical significance:

Spontaneous innovation. In the area with private property, one player accidentally invented what NullCase called "barter bonds" -- forgery-proof tickets for smelting services that functioned as a primitive financial instrument. This was not designed by the server operator; it emerged from players solving real problems within the rule system.

Private commons. Property owners voluntarily created public spaces -- Decentral Park, the Oak Hills commons -- that functioned like urban parks, increasing surrounding property values exactly as they do in the real world. NullCase called this an "invention that came out of Liberty Minecraft."

Transit networks. Players built a world-wide rail system spanning at least 25 kilometers of connected track, with stations owned by more than a dozen different players. The infrastructure was privately funded, privately maintained, and periodically traded on the open market.

Free-market money. A diamond-based monetary system produced purchasing power increases of 10-100x per year, with the money supply expanding at a disinflationary pace through decentralized mining activity.

Predictions from Digital Economies

NullCase extended the chartered-city logic into the future, using observations from Liberty Minecraft to make predictions about the impact of artificial intelligence on real economies. His seven predictions, offered with the caveat that he was not a professional economist, drew directly from watching automation transform the server's economy:

Games offer a low risk, low cost way to explore the impact of artificial intelligence on a digital community. Our AI future and what it means are frontier questions, but they also hit home. Games have a tiny role to play because we can test what happens when people are displaced by automation.

Among his findings: automated production reliably cut resource prices by over 95% within five years. Wages fell but purchasing power outpaced the decline. Manual laborers were still better off one year after full automation because goods became so much cheaper. And Universal Basic Income, which Liberty Minecraft had inadvertently tested during its early years through a technical constraint that paid players $400 per hour of play time, proved counterproductive:

This encouraged idleness. Liberty Minecraft failed twice as a project during this time. People stood around doing nothing to get paid. When UBI was eliminated people learned new skills or invested in automation which produced goods.

These observations were not definitive proof of anything about the real world. But they were data points -- generated at negligible cost, under controlled conditions -- of the kind that cannot be ethically produced through conventional policy experiments.

The Central Authority Problem

NullCase was candid about the fundamental limitation of Minecraft as a governance testbed: every server has an owner with absolute authority. This made true decentralization impossible within the existing technical framework.

All Minecraft servers are owned by someone. The owner has the central, arbitrary authority to decide who can use their server and how. This is a problem for players.

He compared his own position to Satoshi Nakamoto's creation of Bitcoin's genesis block -- an arbitrary starting condition that does not itself constitute central authority, because the system built on top of it is both voluntary and decentralized. But NullCase acknowledged the analogy was imperfect. Unlike Bitcoin, a Minecraft server operator retains ongoing arbitrary power.

The reason Minecraft Servers have a central authority is because one individual has the authority to make arbitrary decisions, and quite often this is true at several levels. I expect that solving this problem is hard, expensive, and worth the risk.

His proposed solution was to decentralize the server itself -- to build a system where no single individual could exert arbitrary authority over players' property. He created OpenLMC, a freely distributable version of Liberty Minecraft, as a first step. But the deeper goal was more ambitious:

The goal would be to put myself out of business by once again providing a better option.

From Game to Laboratory

The chartered-city vision was ultimately about taking games seriously as laboratories for governance. NullCase was not arguing that Minecraft economies perfectly replicate real ones. He was arguing that the cost-to-insight ratio was extraordinarily favorable, that the risks were nearly zero, and that the alternative -- testing governance theories on real populations -- was both expensive and ethically fraught.

Unlike government funded research, a year long experiment can be fully funded for about $1000. This would pay for server costs and some development work. All trade activity can easily be logged on both worlds.

Whether anyone would take up this cause at scale remained to be seen. But Liberty Minecraft had already demonstrated, at the scale of a small community over five years, that a sandbox game with the right charter could produce genuine economic and social phenomena worth studying.

Gallery