Latest / The Productivity Podcast with Fexingo: Output, Efficiency, and Long-Term Economic Growth / How the Smithsonian Solved Its Own Productivity Paradox
Transcript
- Lucas: If you walked through the US economy over the last thirty years, you'd see a weird pattern. Manufacturing got wildly more efficient — robotics, lean systems, global supply chains. But the knowledge sector? Museums, universities, research libraries, archives — those organizations seemed to defy productivity entirely. Luna: The classic Solow paradox — you see computers everywhere except in the productivity statistics. But museums feel like the extreme case. How do you measure the output of a curator? Lucas: Right, and that's exactly the puzzle. So today I want to talk about one institution that actually solved this — the Smithsonian. When COVID hit in 2020, it shut down all 19 of its museums. No visitors, no revenue from gift shops or cafes. And the leadership realized something: if they didn't find a way to produce value digitally, they'd become irrelevant. Luna: That's a pretty existential threat for a 175-year-old institution. But I remember reading that the Smithsonian actually came out of the pandemic stronger — more people engaged with its collections than ever before. Lucas: Exactly. And if today's conversation helps you see how even a giant cultural bureaucracy can rethink productivity, well, that's the kind of thing that keeps this show going. We stay ad-free because listeners who find this useful drop a little support at buy me a coffee dot com slash fexingo. It's a simple way to say 'this mattered to me.' Luna: And honestly, it makes a real difference. We hear from listeners who say these case studies change how they think about their own work. Lucas: Alright — so let's get into the Smithsonian's playbook. The first thing they did was launch what they called the 'Smithsonian Campaign' — a massive digital push. Their goal was to digitize four point two million collection items in eighteen months. That's roughly seven thousand items per week. Luna: Wait — that's an insane pace. How many people did they hire to do that? Lucas: That's the interesting part — they barely hired anyone. Instead, they reallocated existing staff. The Smithsonian has about six thousand employees, and many of them — registrars, curators, conservators — had specialized knowledge of the objects. So instead of outsourcing digitization to a third party, they trained their own people to do it. Luna: So they treated internal talent as a flexible resource, not a fixed cost. That's a pretty radical move for a government-funded institution. Lucas: It is. And they paired that with crowdsourcing. The Smithsonian Transcription Center — originally a small pilot — was scaled up. Volunteers at home transcribed handwritten labels, field notes, and catalog cards. By the end of 2021, they had over ten thousand active volunteers who had transcribed more than a million pages. Luna: That's basically a distributed workforce that costs nothing. And the quality control must have been a challenge. Lucas: They built a review system where each item was transcribed by at least three volunteers, and if they disagreed, a staff member resolved it. The error rate ended up being lower than traditional methods because volunteers were highly motivated — many were retired academics or history buffs. Luna: So the productivity gains came from both sides — internal reallocation and external crowdsourcing. But were the digitized collections actually used? Lucas: That's the key. The Smithsonian made everything available under a Creative Commons Zero license — no restrictions. Anyone could download high-resolution images of objects, use them in textbooks, even put them on T-shirts. Within two years, their online collections had been viewed over two hundred million times. That's more than the total in-person visits in the prior decade. Luna: But does view count equal productivity? I mean, a museum's mission is education and preservation, not just traffic. Lucas: Fair point. But they also saw a measurable shift in educational engagement. Their distance learning programs — virtual field trips, online curricula — went from serving about fifty thousand students a year before the pandemic to over two million by 2023. And they did it with essentially the same education staff. Luna: So output per worker skyrocketed. That's textbook productivity growth. But what about the cost side? Digitization isn't free. Lucas: They spent about thirty million dollars on the digital infrastructure — servers, scanning equipment, software licenses. But they also saved money. By consolidating back-office functions across all nineteen museums — things like HR, finance, IT — they reduced administrative overhead by twelve percent. That's about eighteen million dollars a year. Luna: So the digitization paid for itself in less than two years. And after that, it's pure efficiency gain. Lucas: Exactly. And here's where it gets really interesting from an economics perspective. The Smithsonian created what you could call a 'digital twin' of its physical collections. Now, researchers can study objects without traveling to Washington. A paleontologist in Brazil can examine a fossil from the Museum of Natural History without leaving her lab. Luna: That eliminates a huge travel cost and time delay. It's like the Port of Rotterdam's digital twin, but for cultural heritage. Lucas: Right. And the Smithsonian didn't stop at collections. They applied the same logic to research. Their scientists began publishing open-access papers and sharing raw data. The result? Their citation impact — a measure of how often their research is referenced by other scientists — increased by thirty percent. Luna: Because more people can access and build on their work. That's a knowledge spillover effect — exactly what drives long-term economic growth. Lucas: Precisely. Paul Romer's endogenous growth theory says that ideas are non-rival goods — one person using an idea doesn't reduce its availability to others. The Smithsonian, by making its collections and research openly available, essentially turned its knowledge stock into a public good that fuels further innovation. Luna: But not every museum can do this. The Smithsonian has brand recognition, federal funding, and existing digital infrastructure. What about smaller institutions? Lucas: That's a good challenge. But there are lessons that scale. The Smithsonian used open-source tools like Omeka and CollectiveAccess for their digital collections — free software that any museum can use. And the crowdsourcing model — the Transcription Center — has been replicated by the National Archives and even some university libraries. Luna: So the key insight is about organizational redesign, not just technology. They changed how work was done, not just what tools they used. Lucas: Exactly. And they did something that many companies find hard: they measured output in new ways. Instead of counting visitors through the turnstile, they started tracking digital engagement, research citations, and educational reach. That reframing is what allowed them to see the productivity gains. Luna: So the Solow paradox — 'you see computers everywhere except in the productivity statistics' — maybe the problem isn't the computers, it's the statistics. If you measure the wrong thing, you miss the gains. Lucas: That's exactly right. The Smithsonian's story suggests that productivity in the knowledge sector isn't about doing the same thing faster. It's about doing different things — things that were impossible before digital tools. Freeing up capacity for higher-value activities. Luna: Like a curator who used to spend half her time answering basic inquiries by mail. Now she can focus on researching a new exhibition because the FAQ is handled by an AI chatbot trained on the digitized collections. Lucas: And they actually did that. The Smithsonian deployed a natural language processing system called 'Smithsonian Voice' that answers visitor questions. It handles about eighty percent of inquiries without human intervention. That's a direct labor productivity gain. Luna: So the museum's productivity transformation wasn't just about digitizing artifacts — it was about rethinking the entire workflow from acquisition to public engagement. Lucas: And it's ongoing. In 2025, the Smithsonian launched a new initiative called 'Collections as Data' — they're making the metadata about their collections available in machine-readable formats. Researchers can do large-scale analysis: 'show me all the ceramic objects from the 18th century that were made in Japan.' Pull that up in seconds. Luna: That's a shift from a repository to a research platform. And it makes the Smithsonian more valuable to the economy — not just as a cultural institution, but as a data infrastructure. Lucas: Right. And that's the real productivity story. When you stop thinking of productivity as 'output per hour worked' on a narrow set of tasks, and start thinking of it as 'value created per unit of resource,' you open up entirely new possibilities. Luna: So for any knowledge organization — a university, a law firm, a think tank — the lesson might be: digitize your unique assets, open them up, and then find the new value that emerges. Lucas: And measure that value, even if it's hard. The Smithsonian didn't have a category for 'digital research citations' in 2019. They had to invent it. Luna: I think that's a good note to end on. The productivity paradox might be a measurement problem as much as a performance problem. Lucas: Exactly. And the Smithsonian — a 175-year-old institution with a federal bureaucracy — managed to crack it. If they can, there's hope for the rest of us.