Three Things Aviation Teaches Us About Science Funding

A trip to Long Beach Airport reveals something deep about policy

The LA Uber driver let me off at the small passenger terminal at Long Beach Airport, and I had to do some serious trial and error with Google Maps to find the old Douglas Aircraft hangar where JetZero had set up shop with the admirable goal of completely disrupting the commercial aviation market by building a wide-body blended wing aircraft that would carry a 787 Dreamliner load of passengers across the country for half the fuel cost.

The hangar was open to the air, the ramp and runway fully active, yet the ethos inside was pure early-2000s Google—when anything seemed possible. The enormous space was filled with a full-size cabin mock-up, engineers at workstations, cinema-size screens streaming CAD imagery of the new plane sporting various well-known airline liveries, and a collection of flying scale model drones. The plane itself looked like it had flown off a science fiction set.

The engineering team was equally striking: veterans from Boeing, Embraer, and McDonnell Douglas, each bringing decades of experience from very different aviation cultures. I met one of the chief designers—the inventor of sharklets, those forked wingtips that reduce drag and improve fuel efficiency, now ubiquitous on some commercial aircraft. Another engineer had come from Embraer, where he’d designed the popular 2x2 cabin configuration that passengers overwhelmingly prefer on narrow-body aircraft. Now he was tackling the challenge of designing a completely new kind of airplane cabin that would maximize comfort in a blended wing configuration.

These engineers had learned their craft in established organizations with very different approaches to decision-making, risk assessment, and innovation. The wave of consolidation in aviation—most notably Boeing’s merger with McDonnell Douglas and its subsequent shift from an engineer-driven culture to one focused on shareholder returns—had left many veteran engineers looking for something different. The 737 MAX crisis highlighted how far Boeing had drifted from its engineering roots. JetZero represented a chance to get back to what they loved: solving hard technical problems without the constraints of quarterly earnings calls and legacy infrastructure.

They were attempting something none of their former employers would touch: a radical departure from the tube-and-wing design that has dominated commercial aviation for seventy years. This raised a question that goes far beyond aircraft design: Why can radical innovation happen at a startup like JetZero but not at Boeing, Airbus, or Embraer?

This isn’t just about airplanes. It’s about how organizations—whether aircraft manufacturers or science funding agencies—decide what’s worth building, who gets to decide, and how they balance proven approaches against risky bets. Aviation and science funding face the same fundamental challenge: how to organize technical innovation.

Studying how Boeing, Airbus, and Embraer make these decisions has revealed patterns that apply directly to science funding. Here are three lessons from aviation that illuminate how research gets funded—and why some innovations happen while others never get off the ground.

Lesson 1: How Organizations Assess and Manage Technical Risk

The Aviation Pattern

Boeing, in its traditional engineer-driven culture, approached risk through data and testing. Engineers made decisions based on technical feasibility. They’d prove something worked, then seek regulatory approval. The 787 Dreamliner exemplified this: Boeing pushed carbon-composite technology to unprecedented levels while keeping the basic configuration conventional. The cultural assumption: engineers know best, prove it works, get approval, move forward.

Airbus operates from a completely different framework. As a consortium involving multiple governments, labor unions, and industry stakeholders, risk assessment includes political, economic, and social factors alongside technical ones. Workers’ councils have a voice in production decisions. Safety regulators participate earlier in the design process. The A380 Superjumbo was technically conservative—four engines, conventional configuration—but represented enormous manufacturing and political risks, requiring coordination across nations. The cultural assumption: technical decisions affect many stakeholders, and all deserve input.

Embraer’s approach reflects its position as a state development tool for Brazil (the country holds a veto over control of the company’s strategic direction). They can’t compete head-to-head with Boeing and Airbus, so their risk calculus focuses on market positioning. Find niches, develop partnerships, move quickly. The E-Jet family succeeded by targeting the underserved regional market. The cultural assumption: innovation means finding white space in a market dominated by established players.

Same engineering principles. Same physics. The same goal of building safe, efficient aircraft. But fundamentally different risk assessment frameworks.

The Parallel to Science Funding

The American system, through NSF and NIH, operates remarkably like Boeing’s traditional approach. Peer review is engineer-driven decision-making translated to science. Data—preliminary results, track record—drives decisions. The central question reviewers ask is Boeing’s question: “Can this PI deliver with taxpayer money?” Merit review happens after the proposal is submitted. The system rewards incremental progress from established investigators, just as Boeing refined the 737 through successive iterations.

European research funding embeds more stakeholder involvement. Horizon Europe’s missions approach brings policymakers, industry representatives, and public voices into the priority-setting process. Risk assessment explicitly includes societal benefit and economic impact. Clinical translation gets emphasized earlier in the research pipeline. Scientists remain central but aren’t the sole decision-makers.

Emerging science powers like China take yet another approach. Strategic national priorities drive funding decisions. The question isn’t “What’s the best science?” but “Where can we compete globally?” This enables leapfrog strategies: massive focused investments in AI, quantum computing, and biotechnology designed to establish leadership in emerging fields rather than catching up in established ones. This top-down approach is now also emerging within the US science ecosystem.

For researchers, understanding which risk framework you’re operating in helps you frame proposals effectively. The American system rewards demonstrated competence and incremental progress. Other systems may value societal impact, strategic positioning, or rapid deployment. Neither approach is better or worse—they reflect different cultural assumptions about how to allocate risk in technical innovation.

Lesson 2: Who Gets to Decide What Gets Built

The Aviation Pattern

At Boeing, engineers and program managers traditionally drove major decisions. Shareholders and the board provided financial constraints. Airlines shaped requirements. But core technical choices were the engineers' responsibility. This produced technically sophisticated aircraft, sometimes disconnected from market realities. The 747-8 (the last of the classic jumbo jets’ instantiations), for instance, was an engineer’s dream—but the market for it was lukewarm.

Airbus engages multiple stakeholders from day one. National governments in France, Germany, the UK, and Spain have seats at the table. Workers’ councils negotiate production methods. Industry partners across Europe collaborate on components. Customers get involved earlier. The result is more consensus-driven and sometimes slower, but with broader buy-in. The A350’s long development process reflected extensive consultation but yielded strong market acceptance.

Embraer’s alignment with Brazil’s government development goals sets direction, but the company maintains a partnership model with established players and responds quickly to market signals. Less hierarchical decision-making enables nimble adaptation. The attempted Embraer-Boeing partnership that ultimately fell apart illustrated starkly different decision-making speeds between the two companies.

JetZero represents something different entirely. A small team iterates rapidly. Engineers from different aviation cultures bring different assumptions. Venture capital’s risk tolerance differs fundamentally from corporate risk aversion. They can attempt radical innovation precisely because they’re not constrained by established stakeholder expectations or legacy infrastructure.

The Parallel to Science Funding

American peer review puts scientists in the decision-making seat. On its face, this seems ideal: who better to judge scientific merit than other scientists? But peer review favors known researchers using proven methods. Peers can become conservative gatekeepers. The result is high quality and incremental progress, but potentially missed breakthroughs.

European models bring more voices into the room. The European Research Council maintains scientific independence but operates within frameworks emphasizing societal missions and grand challenges. Policymakers, industry representatives, and public stakeholders help set priorities. Scientists remain central but aren’t the sole arbiters. This creates stronger connections to societal needs, though sometimes at the cost of researcher autonomy.

Directed research models flip the equation. Governments or funding agencies set priorities; researchers respond to calls for proposals. This is top-down rather than bottom-up. The advantage is alignment with national priorities. The risk is missing unexpected discoveries that don’t fit predetermined categories.

I’ve seen these differences firsthand, reviewing for both American and international funding agencies. The questions panels ask reveal cultural assumptions about whose judgment matters. American panels debate scientific rigor and PI capability. International panels I’ve participated in spend more time on broadening participation and strategic fit with national priorities.

For researchers, understanding who has a voice in funding decisions is crucial for navigating the system. American researchers working internationally need to recognize that peer review isn’t universal—other countries organize scientific decision-making to reflect different values about expertise, accountability, and public benefit.

Lesson 3: The Tension Between Incremental Improvement and Radical Innovation

The Aviation Pattern

Established aircraft manufacturers favor incremental improvement for sound reasons. The tube-and-wing design has been refined for seventy years. Every iteration builds on accumulated knowledge. Existing manufacturing facilities, pilot training programs, maintenance infrastructure, and regulatory pathways all assume this configuration. Airlines understand the operating economics. Risk is manageable, returns are predictable. The 737 MAX—an incremental update to a 1960s design—still makes economic sense despite its troubles.

JetZero’s blended wing body has been studied since the 1940s. Its technical advantages are clear: dramatic improvements in fuel efficiency, reduced noise, and potential for entirely new cabin configurations. But it requires new manufacturing processes, new pilot training, and new regulatory frameworks. The risk isn’t primarily technical—it’s organizational and systemic. There’s no clear path from prototype to profitable, scalable production. Established players, accountable to shareholders and constrained by quarterly earnings expectations, can’t justify the investment.

Startups like JetZero can attempt radical innovation because they have no legacy infrastructure to protect. They can accept higher technical risk. The venture capital model tolerates failure in ways public corporations cannot. They don’t need to satisfy existing stakeholders or worry about cannibalizing current product lines. They can focus on long-term disruption rather than next quarter’s earnings.

But we should be clear: most aviation innovation is incremental for good reason. Lives depend on safety. Capital requirements are enormous. Development timelines span 10-15 years. Regulatory burden is intense. Incremental improvement has delivered extraordinary gains—modern aircraft are unimaginably more efficient, safe, and capable than those of fifty years ago.

The Parallel to Science Funding

Science funding faces the same tension. Established PIs using proven methods dominate for sound reasons. Track records reduce risk. Incremental progress is predictable, publishable, and fundable. Infrastructure investments favor established approaches—if your university has a state-of-the-art imaging facility, proposals that use it have an advantage. Peer reviewers understand and can evaluate proven methods. The “preliminary data” requirement inherently favors ongoing work over genuinely new directions. The system is designed to minimize taxpayer waste through careful risk management.

Truly novel approaches struggle in this environment. High-risk/high-reward programs exist but represent a tiny fraction of overall funding. Early career investigators face a chicken-and-egg problem: “How will you do this?” reviewers ask, but gathering preliminary data requires resources they don’t yet have. Reviewers are more comfortable funding known quantities. Paradigm shifts are rare and unpredictable—there’s no clear “return on investment” for genuinely radical ideas.

Consider the BRAIN Initiative. The vision was bold: transform neuroscience through new technologies and approaches. But implementation favored established neuroscientists with proven track records. The system worked as designed: minimizing risk by funding demonstrated competence. As I’ve written earlier, BRAIN fell short in its delivery goals: curing brain diseases. ARPA-H was explicitly created to escape the incremental trap, but it’s still finding its model. The European Research Council’s advanced grants show somewhat higher tolerance for risk, but even there, track record matters enormously.

For researchers pursuing truly novel approaches, it’s crucial to understand you’re working against system design, not just reviewer bias. The system is optimized for reliable incremental progress, not moonshots. Radical innovation in science, like radical innovation in aviation, may require different funding models—something more like venture capital, tolerant of high failure rates in pursuit of occasional transformative breakthroughs.

This raises a deeper question: Should science funding favor incremental or radical innovation? Or do we need both, in different proportions? Aviation supports both Boeing’s incremental refinements and JetZero’s radical rethinking. Should science funding do the same—and if so, in what balance?

What This Means for Science Policy

These aviation patterns reveal a fundamental feature of how societies organize technical innovation. The choices Boeing, Airbus, and Embraer make about risk assessment, decision-making authority, and the balance between incremental and radical innovation aren’t purely business decisions. They’re cultural choices embedded in what Sheila Jasanoff calls civic epistemologies—different assumptions about how knowledge should be produced, who should decide, and what goals matter most.

American science funding has historically reflected American cultural values: individual merit and achievement drive peer review by scientific peers. Data-driven decision-making shows up in preliminary data requirements. Risk minimization operates through proven track records. Incremental progress represents the reliable path. This isn’t accidental—it’s deeply cultural.

Other countries organize differently because they value different things. European systems emphasize societal benefit and stakeholder input. Asian systems prioritize strategic national development goals. Different countries strike different balances between discovery and application, between researcher autonomy and national priorities, between tolerance for failure and demands for accountability.

For all researchers, understanding these cultural patterns helps you work more effectively within the system. Know what the system optimizes for—reliable incremental progress from established investigators. If you’re pursuing radical innovation, recognize you’re working against the grain. International collaborations require understanding that your partners may operate within fundamentally different funding cultures with different assumptions about what science is for and how it should be organized.

For science policy, we should be explicit about what our funding systems optimize for. There’s no “best” system—only different tradeoffs reflecting different values. Maybe we need multiple models, as aviation has both Boeing and JetZero. Comparing systems reveals assumptions we don’t normally question.

In future posts, I’ll explore specific country comparisons: How does the European Research Council actually work? What can we learn from how other countries fund AI research? How do different countries handle the tension between researcher autonomy and national priorities?

A Final Thought

Visiting JetZero and seeing engineers from Boeing, Embraer, and McDonnell Douglas collaborate on something radical that couldn’t happen within their former companies crystallized something I’d been observing in science policy work: innovation doesn’t just require good ideas and talented people. It requires organizational structures and cultural assumptions that allow certain kinds of ideas to be pursued.

The JetZero engineers didn’t suddenly become more creative or capable. They remained the same engineers who’d designed sharklets at Boeing or cabin configurations at Embraer. What changed was the organizational context—the risk tolerance, decision-making authority, and freedom from legacy constraints. That shift in context enabled them to attempt what had been impossible in their former roles.

Science funding works the same way. Researchers operating within NSF’s peer review system are no less creative than those pursuing radical ideas through ARPA or venture-backed biotechs. But the organizational context shapes which ideas can be pursued and which innovations are possible.

Understanding how different countries organize technical innovation—whether building aircraft or funding research—helps us see our own system more clearly. And maybe, just maybe, it helps us imagine how we might do things differently.

What examples have you seen where organizational culture shaped what research got pursued? Have you experienced different funding cultures working internationally? Share in the comments.