When Mahabir Pun demands Nepal spend 2% of GDP on research and development, he channels a nation tired of importing ideas and exporting talent. His frustration is understandable. His prescription is backwards.

The call to match rich nations' R&D ratios rests on a misunderstanding of what those numbers mean. Correlation is not causation. When Israel spends 6% of GDP on research, South Korea nearly 5, and America above 3, it is not the government writing the cheques. In Israel, over 90% of R&D is private-funded by firms racing to commercialise discoveries. Governments fund the seedbed: universities, basic science, and early-stage grants. The real action happens in company labs where competition turns knowledge into profit.

Pun himself proves this point. His wireless networking projects succeeded not through frontier research but by adapting existing technology to solve local problems like connecting remote villages to information and markets. His own work shows that innovation in developing economies comes from applied solutions, not from mimicking rich nations' research budgets.

This distinction matters. Public R&D like NASA or a university's physics department creates knowledge that no one firm can own. Private R&D is what converts knowledge into products: new chips, new drugs, new algorithms. America's vast R&D spending today is driven not by Washington but by the corporate race in artificial intelligence and biotechnology. Microsoft, Google, Meta, and Amazon each spend tens of billions a year on research to perfect AI models. Pharmaceutical giants pour billions into discovering the next blockbuster drug.

These are companies at the frontier, where an additional insight can yield billions in market value.

That is why R&D spending makes sense there. Firms spend because they can capture the returns.

Nepal has none of this structure. It has no large firms with dedicated research arms, no venture-capital networks to fund risky ideas, and almost non-existent university research in applied science. If the country decreed a 2% R&D quota (~ $1 billion annually) it would fall almost entirely on the state. Without capable institutions or commercial partners, that money would generate idle laboratories, inflated conferences, and paperwork, not patents.

This is not cynicism; it is sequence. R&D spending delivers returns only at the frontier. Far from that frontier, the biggest economic gains come not from discovery but from diffusion; adopting, adapting, and improving what already exists. Nepal's economy is nowhere near the frontier in most sectors. Its marginal rupee will yield more by building better factories, not better labs.

Consider South Korea's path. In the 1960s, it spent less than 0.5% of GDP on R&D. Its companies first mastered imitation; assembling radios, cars, and ships based on imported designs. By the 1990s, as firms like Samsung and Hyundai gained technical capability, R&D spending reached 2%. Only after decades of industrial learning did they develop proprietary technologies. Today, Korea spends nearly 5% because it finally has firms capable of turning R&D into revenue. The spending followed capability, not the other way around.

The United States, meanwhile, built its innovation engine through alignment, not allocation. The government funded the basic science through the military, universities, and grants, but private firms turned those discoveries into industries. The internet, GPS, and modern aviation all began with public funding but were scaled by private initiative. The model works because both sides exist.

Nepal lacks both ends of that chain. Its universities do little applied science, and its private sector is too small to commercialise discoveries. Under such conditions, state-led R&D subsidises bureaucracy, not breakthroughs.

The relevant question, then, is not 'how much' Nepal spends on R&D but what kind it pursues. For a developing economy, the priority should be applied research: projects that solve local problems and generate visible payoffs. Research on crop yields, solar energy storage, or digital governance offers immediate utility and can foster collaboration between universities and firms.

Consider agriculture. A small investment in applied soil research could help farmers cut fertiliser use and raise productivity; a return far greater than any theoretical physics lab could offer. In digital technology, focus on software and service innovation, where entry costs are low and local needs are distinct. The success of fintech in Kenya or digital identity systems in India shows that innovation can thrive far from the frontier, provided it solves local pain points.

Nepal's past efforts at state-led innovation offer cautionary lessons. Grand plans for research centres and technology funds have often dissolved into committees and proposals without built infrastructure or sustained partnerships.

The goal, therefore, is not to mimic the spending ratios of rich nations but to recreate the conditions that made those ratios meaningful. Before a country can innovate, it must industrialise; before it can industrialise, it must learn.

Pun's vision of a research-driven Nepal is noble, but his formula risks confusing output with outcome. A nation cannot legislate its way to innovation any more than it can decree prosperity. Until Nepal builds firms that compete, universities that collaborate, and markets that reward risk, research spending will be just another budget line.

The road to innovation runs through enterprise, not appropriation. Before spending billions to invent the future, Nepal must learn to improve the present.

Kharel is a Chartered Financial Analyst and an MBA graduate from University from Toronto