Latest / The Productivity Podcast with Fexingo: Output, Efficiency, and Long-Term Economic Growth / The Productivity Paradox of Open Source Software
Transcript
- Lucas: Luna, quick question. When you think about the software that runs the modern economy, what comes to mind? Windows, maybe, or Oracle databases? Luna: Probably the big proprietary names, honestly. But I know there is a whole layer underneath that most people never see. Lucas: Right. And that invisible layer is open source. Think Linux servers, the Apache web server, the Python libraries that data scientists rely on. A 2026 working paper from the National Bureau of Economic Research tried to put a dollar figure on it. Their estimate: open source software contributed around nine trillion dollars to global GDP in 2025. Luna: Nine trillion. That is roughly the size of Japan's entire economy. How do you even begin to measure something that is free? Lucas: That is exactly the puzzle. And if today's episode makes that puzzle a little clearer, well, that is the kind of thing that keeps this show going. If you get something useful out of it, the way these stays ad-free is listener support — buy me a coffee dot com slash fexingo. Luna: Yeah, we love making these deep dives for you. And a coffee or two helps keep the lights on. Lucas: So, back to the nine trillion. The paper's authors — they are economists at Harvard and MIT, among others — they did not just count the market value of open source firms. That would miss the point completely. Luna: Because the value is not in selling the software, it is in using it. Every company that runs on Linux is getting that productivity for free. Lucas: Exactly. So they took a different approach. They measured the 'consumer surplus' — the value users get beyond what they pay, which is zero. They estimated how much it would cost a company to recreate the open source software they use internally, and then they added the value of the innovation that open source enables. Luna: Recreating it internally. That is a key point. If you had to build your own operating system or your own database from scratch, the cost would be astronomical. Lucas: Right. And the paper's number — nine trillion — is about 7 percent of global GDP. That is bigger than the contribution of many entire industries. But here is the twist: because open source is free, it is almost invisible in standard productivity statistics. Luna: So the GDP figures we rely on are missing a massive chunk of value creation. That seems like a problem for policymakers. Lucas: It is. And it is not just about measurement. It is about incentives. If open source is that valuable, who maintains it? A lot of it is maintained by volunteers or by companies that use it, but there is a real sustainability issue. Luna: We saw that with log4j, right? That was a critical logging library used everywhere, and it was maintained by a tiny group of volunteers. When a vulnerability hit, the whole internet scrambled. Lucas: Exactly. Log4j is a perfect case study. It is a small piece of Java code that runs in millions of applications. The maintenance burden falls on a few people. The NBER paper actually tries to quantify the cost of that maintenance — they estimate the total development cost of open source software is somewhere in the hundreds of billions of dollars. Luna: Hundreds of billions. And yet we treat it as if it cost nothing. That is a huge market failure. Lucas: It is. And the paper suggests that if we do not address it, we risk losing some of that productivity growth. They propose policy ideas like a 'digital public goods' fund or tax incentives for companies that contribute back to open source. Luna: Interesting. So the idea is to make the invisible infrastructure more sustainable. But I wonder, how do companies think about this? IBM bought Red Hat for thirty-four billion dollars in 2019 — that is a huge bet on open source. Lucas: Right. And Microsoft, which used to call Linux a cancer, now is one of the biggest contributors to open source. They have realized that embracing it is better for business. The NBER paper highlights how companies like these use open source to drive innovation in their proprietary offerings. Luna: So it is not just altruism. There is a business case. You contribute to open source to shape the direction of the technology, and then you sell services or cloud access around it. Lucas: Precisely. The paper notes that the majority of open source contributions now come from employees at tech companies, not hobbyists. In 2025, around 80 percent of contributions to major open source projects came from corporate developers. Luna: That is a shift from the early days. And it makes the measurement problem even more complex, because the value is embedded in proprietary products too. Lucas: Yes. And that is why the nine trillion figure is actually conservative, in a way. It only counts direct use value, not the knock-on effects. For example, a startup might use open source to build a product that becomes a unicorn — that value is not captured in their estimate. Luna: So the real contribution could be even higher. But if we cannot measure it, we cannot manage it. That is the heart of the productivity paradox. Lucas: Exactly. And this is not just an academic issue. The paper argues that because open source is not counted in GDP, we underinvest in it. Governments and companies allocate resources based on where they see returns, and open source looks like it returns nothing. Luna: But we know it returns a lot. So how do we fix that? Do we need to change the way we measure GDP? Lucas: That is a big question. Some economists, like the paper's authors, suggest we need to develop better satellite accounts — like we do for research and development — to track open source value separately. That would at least give policymakers a clearer picture. Luna: And then maybe we could make better decisions about funding critical infrastructure. Like the open source equivalent of roads and bridges. Lucas: Right. And there is some movement on that. The European Union, for example, has funded security audits for open source software. And the White House has talked about securing open source supply chains. But it is slow going. Luna: So we are starting to see the policy response, but it is still dwarfed by the scale of the problem. A nine trillion dollar asset deserves more than a few grants here and there. Lucas: Absolutely. And that is the kind of thing the paper is trying to wake people up to. They are not saying open source is going to collapse, but they are pointing out that the current model relies on a few overworked maintainers and a lot of corporate goodwill. Luna: And that goodwill could dry up if the economy slows down. Companies might start cutting their open source budgets. Lucas: That is a real risk. And the paper actually models what happens if contributions drop — productivity growth could take a hit. They estimate that a 10 percent reduction in open source investment could shave off a couple of tenths of a percentage point from GDP growth. Luna: That does not sound huge, but over time it compounds. And in a world where productivity growth is already slow, we cannot afford to lose even that. Lucas: Exactly. So the bottom line is, open source is the invisible engine of the digital economy. It is free, it is everywhere, and it is worth trillions. But we treat it like a public park — we love it, but nobody wants to pay for the maintenance. Luna: That is a great way to put it. So next time you use a open source library, maybe think about the people who built it. And maybe see if there is a way to contribute back, even if it is just a donation. Lucas: Or a pull request. But seriously, the NBER paper is a wake-up call. We need to start measuring what matters, or we risk losing it. Thanks for listening, everyone.