Latest / The Productivity Podcast with Fexingo: Output, Efficiency, and Long-Term Economic Growth / The Long Tail of Productivity-Boosting Capital Investment
Transcript
- Lucas: We talk a lot on this show about the micro habits, the office layouts, the meeting cultures that supposedly boost or kill productivity. But there's a much bigger, slower-moving force that doesn't get nearly enough airtime. Luna: Let me guess — you're about to say capital investment. Lucas: Exactly. But not just any capital investment. I want to talk about the specific, measurable delay between when a company buys a piece of equipment or installs new software and when that investment actually shows up in productivity data. And the numbers from a recent Federal Reserve study are pretty striking. Luna: Alright, I'm hooked. How long are we talking? Lucas: So the San Francisco Fed published a working paper in 2025, tracking the productivity impact of capital spending across US industries over a 25-year window. Their headline finding: the median investment takes twelve to fifteen years to deliver its full productivity benefit. Luna: Wait — that's not just the payback period. That's the time to reach peak productivity contribution? Lucas: Correct. The equipment or software itself might be installed in a year, but the productivity gains keep accumulating for over a decade. The study calls this the 'long tail' of capital investment. And it's largely driven by complementarities — the training, the process redesign, the organizational changes that need to happen before the new machine or system actually changes how work gets done. Luna: So it's not the machine that makes you more productive — it's everything you do around the machine over the next ten years. Lucas: Exactly. And the study found that about 60 percent of the total productivity benefit from a given investment arrives more than five years after the purchase. That's a huge mismatch with how most companies evaluate capital projects, where the standard net present value model heavily discounts returns beyond three to five years. Luna: That creates a real tension. If your CFO is using a three-year payback threshold, you're systematically rejecting investments that would pay off massively in years eight through fifteen. Lucas: Right. And the study didn't just look at physical equipment like factory robots or MRI machines. They also looked at software and IT systems, and the long tail was actually more pronounced there. Software investments took even longer because the organizational change — new workflows, data standards, employee training — is more complex. Luna: I can see that. My old company implemented a new CRM system, and for the first two years, productivity actually dropped. The sales team hated it. It took about four years before we saw the promised efficiency gains. Lucas: That anecdote maps perfectly onto the data. The study's authors call that initial dip the 'productivity J-curve' — output falls before it rises. And the fall can be significant. For software investments, the average productivity decline in years one and two was about 8 percent below baseline before things turned positive. Luna: Eight percent is brutal. Especially when quarterly earnings calls are punishing any sign of weakness. Lucas: And that's the core tension. Public company CEOs face enormous pressure to show results within a few quarters. But the productivity-boosting investments that actually move the needle on long-run growth take more than a decade to mature. The study basically argues that the way we measure and incentivize capital allocation is structurally biased against the most productive investments. Luna: Does the study offer any policy solutions? Lucas: It does. The authors suggest that tax incentives — like accelerated depreciation or R&D tax credits — should be structured to reward long holding periods rather than just the initial spend. They also recommend that companies report internally on 'productivity investment maturity' — essentially tracking the age distribution of their capital stock so they can see when past investments are starting to pay off. Luna: Okay, but isn't there a risk that this becomes an excuse for bad investments? 'Don't worry, this terrible ROI will look great in 2038'? Lucas: That's a fair challenge. And the study addresses it by distinguishing between investments that show leading indicators — early improvements in quality, customer satisfaction, employee skill acquisition — and those that don't. The long tail argument only works if you have evidence that the complementary investments are actually happening. If you buy a factory robot and never retrain your workers, you won't see the tail. Luna: So what's the practical takeaway for a business leader listening right now? Lucas: I'd say two things. One: when you evaluate a capital project, build in a realistic J-curve. Plan for a productivity dip in the first two years and set expectations accordingly. Two: track the age of your capital stock and the maturity of your past investments. If you have a lot of spending from seven years ago, you might be entering the golden period of productivity payoff — and you should double down on those complementary investments rather than abandoning them for the shiny new thing. Luna: And for policymakers? Lucas: The study makes a strong case for tax structures that reward patient capital. One specific proposal: allow companies to carry forward net operating losses from productivity investments for longer periods — maybe twenty years instead of the current limit. That would directly reduce the penalty for the early years of the J-curve. Luna: That's interesting. It aligns with something we've talked about before — the idea that short-termism is baked into our accounting and tax rules, not just our culture. Lucas: Exactly. And this study gives us a concrete number to attach to that idea. The median productivity peak at twelve to fifteen years is a powerful anchor. It says: if you want to grow productivity in the long run, you need to invest for a decade and a half, not a year and a half. Luna: And honestly, that's a hard sell in a world where the average CEO tenure is about five years. Lucas: Right. The CEO who makes the investment won't be there to reap the productivity glory. That's a structural misalignment that no single company can solve on its own. But being aware of it probably changes how you advocate for budget within your own organization. Luna: Yeah, I think that's the real value of the study — it gives you ammunition to argue for patient capital inside your own firm. Lucas: One thing I really appreciate about this paper is that it doesn't just say 'invest more.' It says 'invest with a fifteen-year horizon in mind, and measure the complementary stuff.' That's actionable. Luna: On complementary stuff — the study mentions organizational capital. Can you unpack that? Lucas: Sure. Organizational capital is the internal knowledge, processes, and culture that makes physical capital productive. The study estimates that for every dollar spent on equipment, you need to spend about forty cents on organizational capital to realize the full productivity benefit. And that forty cents is spent over the following decade, not upfront. Luna: So when a company buys a million-dollar machine, the real cost is actually one point four million, but the extra four hundred thousand is spread out over ten years and often not tracked as a capital expense. Lucas: Exactly. And because it's not tracked as capital, it gets cut in budget tightening. That's how companies sabotage their own investments. They buy the machine but starve the organizational change it needs. Luna: So how do you fix that? Do you force companies to capitalize training costs? Lucas: The study suggests that, or at least create a separate line item for 'organizational capital expenditure.' Some companies are starting to do this internally — they call it 'change management budget' or 'implementation spend.' But it's still rare. Luna: Let's talk about one example that might make this concrete. The semiconductor industry. Extreme capital intensity, very long product cycles. Lucas: Great example. A chip fabrication plant costs ten billion dollars plus. The productivity payoff from that plant is measured over fifteen to twenty years. And the companies that do it well — like TSMC — are the ones that invest continuously in the process improvements and training over that entire period. They don't just flip a switch. Luna: And they're also the most productive semiconductor companies. Lucas: Correlation, not necessarily causation, but it's suggestive. The study's framework would say: TSMC's productivity is not just from the equipment, it's from two decades of organizational capital accumulation around that equipment. Luna: That's a powerful lens for understanding why some companies seem to pull away from competitors despite similar capital spending. Lucas: It really is. And it also explains why productivity growth has been so uneven across industries. The industries that invest heavily in organizational capital — like tech and manufacturing — see bigger long-run gains. Others that treat capital as a plug and play solution don't. Luna: I want to push back on one thing. The study is based on historical data from mostly physical capital-intensive industries. Does it apply equally to software and services? Lucas: That's a good question. The study does include software spending, and the authors argue the mechanism is the same. But you're right that the organizational capital component might be even larger in services because the 'equipment' is often just a platform and the real work is human adaptation. They call for more research on that. Luna: Okay, so given all this — what's one thing a listener could do differently starting next week? Lucas: Look at your company's capital spending from seven years ago. Find that batch of investments. Are the complementary organizational changes still being funded? Are you measuring the productivity trajectory? If not, that's probably where the next wave of productivity gains is hiding — and you're leaving it on the table. Luna: That's a very concrete homework assignment. I like it. Lucas: And speaking of homework — you know, we do this show every weekday, and it's entirely listener-supported. We don't run ads, we don't sell data, we don't have sponsors. It's just a handful of people who chip in through buy me a coffee dot com slash fexingo who make it possible. Luna: Yeah, and it's a small group, but it's enough to keep the lights on and keep the research flowing. If you've gotten value from our deep dives into topics like this capital investment lag, that's exactly who we're doing it for. Lucas: So if that's you, buy me a coffee dot com slash fexingo. No pressure, no perks, just knowing you're part of keeping this independent. Luna: Exactly. And back to the topic — I think one of the most underappreciated aspects of this study is its implication for how we teach capital budgeting in business schools. Lucas: Oh, that's a great point. Most MBA programs still teach discounted cash flow with a five- to ten-year horizon. But if the productivity payoff peaks at twelve to fifteen years, that's a fundamental mismatch. Maybe we should be teaching a 'productivity-adjusted payback period' that accounts for the J-curve. Luna: And teaching students to look for organizational capital as a real asset, not just a soft cost. Lucas: Exactly. If we can shift the way managers think about capital investment — from a purchase event to a long-term capability-building process — we could unlock a lot of productivity growth that's currently being left on the table. Luna: And that's the kind of structural change that actually moves the needle on national productivity, not just individual company performance. Lucas: Right. And that's why we spent this episode on it. Because the small stuff matters, but the big stuff — the stuff that takes fifteen years — that's what builds lasting prosperity.