Ethical Technology

96% of Commercial Software Contains Code That Volunteers Wrote for Free

Wikipedia, Linux, OpenStreetMap: three projects built by millions of people with no single owner. Harvard puts their replacement value at $8.8 trillion. The data demolish the idea that open collaboration cannot work at scale.

Published on 11 min de lecture

96% of Commercial Software Contains Code That Volunteers Wrote for Free

TL;DR: Millions of people, with no shared employer and no shared commercial goal, are building together a digital infrastructure the entire world depends on. Wikipedia: 60 million articles, 23 billion views per month. Linux: 100% of the world’s 500 most powerful supercomputers. OpenStreetMap: 10 billion map nodes, used by Apple, Amazon, and Uber. Harvard values the replacement cost of these commons at $8.8 trillion. This is not idealism — it is documented economics.

Imagine a single company had to rebuild, from scratch, all the open-source code that makes the internet run. What would that cost?

Researchers at Harvard posed the question seriously. In 2024, Frank Nagle and colleagues at the Laboratory for Innovation Science scanned 9 million company websites, analysed the structure of the software in use, and applied a standard industry cost model. Their answer: $8.8 trillion. Companies would pay 3.5 times more for their software if open source did not exist. And 96% of commercial programs contain code written, modified, or distributed free of charge by public contributors.

That figure should stop cold anyone who still believes “open collaboration cannot work at scale.” It works. It runs your phone, your bank, your hospital.

What digital commons are

The term comes from economist Yochai Benkler, who coined the concept of commons-based peer production in 2001. His definition: a process in which “inputs and outputs are shared, freely or conditionally, in an institutional form that makes them equally accessible to all”, and where individuals act “from social and psychological motivations — doing something interesting”.

No salary. No owner. No imposed hierarchy. Just governance rules, a shared goal, and thousands of people who find meaning in contributing.

Benkler develops the idea in The Wealth of Networks (Yale University Press, 2006): this mode of production, marginal before the internet, becomes central once coordination costs collapse. The theory was bold. The facts since then have validated it decisively.

The impossible scale: concrete evidence

Wikipedia: humanity’s library

In 2026, Wikipedia has more than 60 million articles in over 300 languages. Each month, servers record 23 billion page views. The English edition alone exceeds 6.7 million articles — roughly 4.4 billion words, about 20 times the length of the Encyclopædia Britannica in its complete edition.

The United Nations has recognised Wikipedia as a «digital public good», highlighting its role in universal access to verified knowledge. Annual budget of the Wikimedia Foundation: around $185 million in revenues for 2023-2024, mostly from donations. Replacement cost: incalculable.

What is striking is not just the scale — it is the reliability. Studies comparing Wikipedia to professional encyclopaedias consistently find similar error rates for mature articles. Peer production, with millions of potential reviewers, converges on accuracy more reliably than production by a small team of experts.

Linux: the invisible engine of digital civilisation

Your Android phone runs on Linux. Amazon, Google, Facebook, and Netflix servers — Linux. The International Space Station — Linux. The 500 most powerful supercomputers in the world — 100% Linux since November 2017. In 2024, Linux controls 44.8% of the server operating system market and runs 49.2% of cloud workloads.

Linus Torvalds launched the project in 1991 with a now-famous announcement: “I’m doing a (free) operating system, just a hobby, won’t be big and professional…” Thirty-five years later, that hobby represents tens of billions of dollars in annual economic value.

In 2026, the Linux Foundation published a study of 500+ technology leaders: organisations that actively contribute to open source achieve a 2 to 5 times return on investment. Between 2018 and 2025, the top 100 open-source contributors generated $23.2 billion in value from a $3.9 billion investment — a 6x multiplier.

OpenStreetMap: the free map of the world

Launched in 2004 in London by Steve Coast, OpenStreetMap is today the most detailed free map ever created. The numbers are staggering: 10 million registered users, 2.25 million active contributors, 10 billion map nodes. Around 4 million edits per day.

Who uses this map? Amazon, Apple, Baidu, Facebook, Microsoft, Uber, Snap, Deutsche Bahn, Air France, the New York City government, the French government for its national address database, and Indonesia for disaster management. Companies with market capitalisations in the hundreds of billions — built in part on the unpaid work of volunteers.

The most striking case is humanitarian. HOT — the Humanitarian OpenStreetMap Team — mobilises 758,000 volunteer mappers across 94 countries. When the 2010 earthquake struck Haiti, when the 2015 earthquake devastated Nepal, HOT deployed its teams to map the affected zones in real time — enabling rescue workers to find passable roads, standing hospitals, and access points. By 2026, 1.02 billion people live in areas covered by HOT’s crisis maps.

The invisible economic value

The paradox of digital commons is this: they produce enormous economic value without capturing it.

Harvard measured it: $8.8 trillion in replacement value — roughly the combined GDP of Germany, Japan, and the United Kingdom. The Linux Foundation confirms it: the return on active contribution is 2 to 5 times. And yet, only 5% of developers create more than 90% of the value — and most work without direct compensation.

This is not a malfunction — it is a structural feature. Digital commons work precisely because they cannot be enclosed. Once a piece of software is free, anyone can use, modify, and improve it. Code does not wear out through use. It costs nothing to duplicate. Classical economic rivalry disappears.

But this property also creates a tension: those who benefit from the commons do not necessarily contribute to its upkeep. Economist Elinor Ostrom, 2009 Nobel Prize winner, spent her career studying how commons avoid over-exploitation. Her eight governance principles — clear boundaries, rules adapted to context, user participation in decision-making, internal monitoring, graduated sanctions — were formulated for natural resources. They apply remarkably well to digital commons.

Honest limits: what can crack

The invisible dependency

In December 2021, a security flaw was discovered in Log4j, a near-universal Java library used to log events in software. Its criticality score: CVSS 10 — the highest possible. Billions of devices, from Fortune 500 servers to Minecraft servers. Potentially affected.

Behind this critical library: a volunteer maintainer. Christian Grobmeier, one of the Log4j developers, recounted: “Some of us stopped sleeping. We all felt that either we fixed this now in a few days, or we’d shut down the project.” He added: “Nobody stops to ask how you’re doing. They check the project’s status. Nobody says, ‘Hey, thank you for the good work you’re doing.’”

Maintainer burnout

This is not isolated. According to the Tidelift 2024 report (via byteiota): 60% of open-source software maintainers are not compensated for their work. 44% cite burnout as the primary reason for abandonment. 60% have left or considered leaving their project.

The XZ Utils affair in 2024 is emblematic. An exhausted maintainer, under constant pressure, accepted help from a stranger who turned out to be a malicious actor who had patiently built a relationship of trust. A backdoor in a compression tool used in almost every Linux distribution. Discovered barely in time by a Microsoft engineer, by chance.

The lesson is not that open source is dangerous — it is that its security is proportional to the care society extends to those who maintain it.

The concentration of contributions

The Harvard study highlights a troubling asymmetry: 5% of developers create more than 90% of the value. The horizontal governance of digital commons does not eliminate hierarchy — it relocates it toward reputation, technical capacity, and availability.

Political fragility

Munich offers the textbook case. Between 2003 and 2012, the city migrated 12,600 workstations to LiMux, a municipal Linux distribution, and saved a documented €11.7 million. In 2017, a change in political majority reversed the decision: back to Windows. The savings were real. The reasons for the reversal were political, not technical.

But the story has an epilogue. In October 2024, the city of Munich reinforced its open-source commitment with a five-point plan and created an Open Source Programme Office. Digital commons are vulnerable to political decisions — and can also benefit from them.

Why it holds when it holds

The question is legitimate: what prevents digital commons from collapsing under the weight of passivity, vandalism, or malice?

Three mechanisms combine.

Governance by rules. Wikipedia works because its contribution rules — neutral point of view, verifiability, no original research — are clear, public, and enforced by the community. Important decisions go through deliberately slow consensus processes. This is not anarchy; it is distributed governance.

Intrinsic motivations. Benkler was right: people who contribute to open source do so primarily for non-monetary reasons — the satisfaction of solving a hard problem, peer recognition, the sense of building something that lasts. These motivations are durable in ways that compensation alone cannot guarantee.

Modularity. Linux and Wikipedia are decomposable architectures: each contributor can work on a module without understanding the whole. This multiplies the number of people who can participate usefully, and creates distributed resilience — if one contributor disappears, others can pick up.

France has begun to institutionalise this model. The government allocated €30 million from the recovery plan to open-source solutions for local authorities. The code.gouv.fr platform brings together more than 9,000 repositories from over 100 public bodies. This is the formal recognition that digital commons are not a hobby — they are public infrastructure.

What this says about us

There is a stubborn received idea: humans are fundamentally competitive, individualistic, short-term strategists. Neoclassical economics made this an axiom. The market made it a design.

Digital commons tell a different story. Not an idyllic one — the burnouts, the vulnerabilities, the political reversals are all too real. But a factual one: when the conditions are right — clear rules, recognition of contributions, modularity, absence of rent capture — millions of people work together over decades to build something that belongs to no one and benefits everyone.

Wikipedia started as a joke — nobody seriously believed an encyclopaedia open to all would survive the vandals. Linux started as a Finnish student’s weekend project. OpenStreetMap started in a London flat because Steve Coast found commercial maps too expensive.

These projects did not succeed because their founders had raised exceptional funding or recruited extraordinary teams. They succeeded because they had solved the governance problem well: how do you make contribution accessible, recognised, and protected from exploitation?

The question is no longer whether open collaboration can work at scale. It does. The question is: what do we want to build together next?

What you can do

  • Contribute to Wikipedia: fix an error, expand an article in your field of expertise. Ten minutes of your knowledge is worth hours for someone searching for it.
  • Use and report issues on OpenStreetMap: every correction improves the crisis maps available at the next disaster.
  • Support maintainers: platforms like GitHub Sponsors allow you to fund directly the developers your daily software depends on. Even €5 a month changes the reality of an unpaid maintainer.
  • Advocate for open-source policies: in France, code.gouv.fr and the DINUM policy show that public institutions can play an active role. Asking your local elected representatives about digital sovereignty is a concrete lever.

Further reading:

Sources

Pass it on

If this article resonated with you, share it with someone around you. In a world where bad news travels fast, it is just as valuable to spread what is going well. Passing on something inspiring is already a form of action.

Leave a comment

Did this article inspire a thought, a question, or a reaction? Share it below — all comments are reviewed before publication.

0 / 1000

Your data is used solely to process this comment and will never be sold or shared with third parties.

Related articles