Why Markets Underfund Basic Research

Getting your Trinity Audio player ready...

Markets tend to underfund basic research because the organization paying for a discovery can capture only part of the value it creates. Benefits spread to other companies, researchers, consumers, and future generations. Research that is worthwhile for society may therefore be unattractive to a private investor.

The central distinction is between creating value and earning revenue from that value. Basic science can succeed at the first while offering little opportunity for the second.

What is basic research?

Basic research investigates fundamental questions about how the world works, without targeting a particular practical application. Applied research pursues a specific practical objective; experimental development uses knowledge to create or improve products and processes. These distinctions follow the OECD’s Frascati Manual.

“Basic” does not mean simple or unimportant. It describes the purpose of the investigation, rather than its difficulty or eventual usefulness.

The gap between private and social returns

A private return is the benefit captured by the investor. A social return includes benefits to others, such as improved health, lower production costs, and knowledge that enables further discoveries.

Consider a hypothetical research project:

Expected outcome, expressed in present value	Amount
Research cost	$10 million
Benefit the funder can capture	$3 million
Additional benefits to everyone else	$47 million
Total social benefit	$50 million

Assume these estimates already account for uncertainty, timing, and relevant harms. Society gains $40 million after costs, but the funder loses $7 million.

Declining the investment can be commercially rational even when undertaking it would be socially beneficial. These figures illustrate the mechanism; they are not empirical estimates.

Underfunding means investing less than is socially worthwhile—not that every proposed research project deserves money.

Why basic research is especially exposed

Knowledge benefits people who did not pay for it

Once shared, a scientific insight can support many users simultaneously. One researcher’s use of a theorem does not prevent another from using it. Economists call this property non-rivalry.

Knowledge can also be difficult to keep exclusive after disclosure. These characteristics create spillovers: benefits that reach people outside the original funding arrangement.

Richard Nelson’s foundational paper, “The Simple Economics of Basic Scientific Research”, explains why the broad applicability of basic science can weaken private incentives to support it. A discovery may serve many industries, while its sponsor operates in only a few.

Future customers and applications may be unknown

A research team may know what question it wants to investigate without knowing which product could eventually result. That makes it difficult to identify customers, estimate revenue, or arrange contracts with future beneficiaries.

Uncertainty alone does not establish market failure: investors routinely fund risky ventures. The stronger problem arises when uncertain discoveries also generate benefits that the investor cannot capture.

Even an extremely patient investor needs some route from future usefulness to future income.

Everyone can prefer that someone else pays

Imagine several companies that would benefit from an openly published scientific result. Each would welcome the research, but each could prefer to let another company fund it.

Joint funding can help. However, negotiating contributions becomes harder when beneficiaries are numerous, dispersed, or not yet identifiable.

This is the free-rider problem: people can have a reason to withhold payment even for something they value.

What does the evidence show?

The economic case goes beyond a hypothetical example.

Charles Jones and John Williams’s Quarterly Journal of Economics article, “Measuring the Social Return to R&D”, connects estimates of research returns to a growth model and finds substantial underinvestment. An accessible working-paper version explains the analysis.

This concerns R&D broadly. It does not establish a universal return for basic research, an exact funding target today, or the value of an individual scientist’s work.

More specific evidence shows how public research can support commercial innovation. Pierre Azoulay and colleagues studied NIH funding and subsequent pharmaceutical and biotechnology patenting. Their published analysis estimates that an additional $10 million in NIH funding produced a net increase of 2.7 private-sector patents. See “Public R&D Investments and Private-sector Patenting”.

Patent counts do not measure the full social value of research, and NIH grants are not exclusively basic research. Nevertheless, the study supports a concrete connection between publicly funded science and later private invention.

Why companies still fund basic research

Underfunding does not imply zero private investment.

A company may benefit from earlier access to discoveries, specialist expertise, recruitment, or applications across several business lines. Research partnerships can also distribute costs among beneficiaries.

These advantages can make some basic research commercially attractive. They do not ensure that every socially valuable field has a profitable business model.

The relevant question is therefore whether private incentives cover the full range of worthwhile research—not whether businesses ever do fundamental science.

Which funding approaches can help?

Different mechanisms address different parts of the problem:

Approach	What it can support	Main limitation
Public funding	Research with broad benefits beyond individual firms	Allocation can suffer from bureaucracy, bias, and political priorities
Philanthropy	Neglected questions and work without commercial returns	Depends on donor resources and preferences
Industry consortia	Shared research useful to participating organizations	May overlook benefits outside the membership
Prizes and advance commitments	Work toward identifiable, verifiable outcomes	Difficult to specify rewards for discoveries nobody can predict
Retrospective rewards	Contributions whose significance becomes clearer after publication	Researchers still need resources before producing results

This is a design comparison, not a claim that one mechanism always performs best. A sensible funding system can combine upfront support with later recognition.

Where AIIM could contribute

AI Internet-Meritocracy (AIIM) is Science DAO’s experimental approach to allocating donated funds using AI-assisted assessments of documented scientific and software contributions.

Its relevance is straightforward: researchers could receive support for openly shared work even when they cannot charge everyone who benefits. The proposed advantage is a funding route based on assessed contributions rather than ownership of a commercial product.

However, allocation and fundraising are separate problems. AIIM cannot eliminate the free-rider problem merely by evaluating contributions. It still needs donors or other funders willing to finance shared benefits.

Retrospective assessment also requires safeguards against errors, prestige bias, manipulation, and the neglect of work whose importance emerges slowly. AIIM’s effectiveness remains a hypothesis to test, rather than an established solution to basic research underfunding.

Funding the foundations of future innovation

The market price of a discovery can be a poor guide to its social importance. Widely reusable knowledge may create benefits that never become revenue for its creator.

That gives society a reason to fund basic research collectively—and to scrutinize how that funding is distributed. Public institutions, philanthropy, industry partnerships, and experimental systems such as AIIM should be judged by whether they help valuable research happen, remain accessible, and support further discovery.