Contacts
How a bold new economic model explains why science sometimes needs to ask riskier questions

Every year, billions are spent on scientific research in the hope of solving real-world problems—from climate change to pandemics. Yet most of that science moves cautiously. Researchers tend to ask questions close to what we already know. Why? Because it’s safer. The odds of success are higher, and rewards—grants, publications, tenure—depend on outcomes.

But what if this very caution is slowing science down?

In A Quest for Knowledge (Econometrica, March 2025), economists Christoph Carnehl (Department of Economics/IGIER, Bocconi) and Johannes Schneider (Department of Economics, Universidad Carlos III - Madrid) offer a striking answer: science tends to evolve through a “ladder”—a series of small, predictable steps. This structure, while efficient in the short term, leaves much of the intellectual landscape unexplored.

“Researchers work on questions that are too narrow and fail too often. The questions chosen are neither novel enough to maximize the instantaneous value of knowledge… nor novel enough to inspire future generations.”

The map of knowledge

To explain how this happens, the authors introduce a model in which knowledge is like a path across unknown terrain. Scientists can choose where to explore, but the farther they go from what’s already known, the harder it is to find answers. Novel research is riskier, and it requires more effort.

So, researchers do what’s rational: they stay close to home. That means safer questions, fewer failures—but also fewer breakthroughs. Over time, knowledge grows, but only in a narrow band. Carnehl and Schneider call this pattern a ladder.

“The evolution of knowledge takes on a ladder structure.”

This helps explain why so much scientific research, even at elite institutions, can feel incremental: because the system unintentionally rewards proximity over originality.

Why moonshots matter

The alternative? Take a leap.

Carnehl and Schneider argue that moonshots—research projects that target questions far beyond the current frontier—can jolt the system. On their own, moonshots may seem disconnected or even impractical. But once made, they change the landscape. They create a new reference point. Other researchers begin to explore the space in between.

“Moonshots have the flavor of an infrastructure investment. By themselves, they provide little value, yet they enable future researchers to thrive.”

This is what the authors call a research cycle. A moonshot triggers not just one discovery, but a cascade. The knowledge that emerges is richer, more interconnected—and more useful.

Designing for discovery

The paper doesn’t argue that every scientist should shoot for the moon. Most won’t—and shouldn’t. But it does suggest that science policy needs to create space for those who will.

When research is moderately expensive and funders are patient, moonshots become worth it. They don’t just expand knowledge; they transform it. But to make that happen, funding institutions need to look beyond short-term productivity. They need to reward vision.

“Focusing only on immediate policy relevance when designing researchers’ incentives is suboptimal for patient societies.”

That means valuing influence over time, supporting bold ideas with uncertain outcomes, and funding the kind of curiosity that doesn’t always fit inside a grant application box.

Carnehl and Schneider offer more than a model. They offer a philosophy: that sometimes the best science isn’t the kind that solves today's problems—it’s the kind that makes tomorrow’s questions possible.

Research