Microeconomics › Section M · Positive externalities and public goods › Topic M.2

How governments encourage innovation

The spillover leak has three plugs in a government's toolbox: exclusive rights that let inventors charge for their ideas, public money that pays for research directly, and partnerships that stitch universities to firms. Each plug has a price - and the art of innovation policy is knowing which plug fits which leak.

Three policies, one leak

The previous topic left the problem: private returns fall short of social returns, so markets underinvest in research. Governments answer with three broad instruments. Intellectual property rights: patents and copyrights that let the innovator capture more of the benefit. Government assistance with research costs: grants, public laboratories and tax breaks that lower the private bill. Cooperative research ventures: partnerships between universities and companies that share costs and let knowledge flow. Every real innovation system mixes all three.

Policy one: intellectual property

The most famous plug: grant the inventor sole legal control of the invention for a limited time - patents for products and processes (roughly two decades), copyright for expression (the author's life plus decades). The rights convert spillover into sale: a pharmaceutical firm with a patent can charge monopoly prices for its drug's protected years, earning returns that repay the research. Without protection, competitors copy immediately and the best any innovator earns is a normal profit - not enough to fund a decade of trials. IP rights do not create the externality; they fence part of it in, trading a period of monopoly pricing for the disclosure of the invention.

The patent bargain's fine print

Patents are a fix, not a cure. Economic studies suggest inventors still capture only about a third to a half of their inventions' total value - the fence is leaky by design. In fast-moving fields like semiconductors or biotechnology, technology may outrun a patent before it matters. Some of the most valuable ideas - a better factory organisation, a training method - are not patentable at all. And patents can be over-broad or over-easily granted, letting firms fence off ground they never ploughed - a weakness 'patent trolls', who hold patents mainly to sue, exploit. The length is arbitrary too: long enough to repay research, short enough to limit monopoly - a balance, not a constant.

Copyright and the extension debate

Copyright lasts far longer than patents - in the US, the author's life plus 70 years, and 95 years from publication for corporate-owned works. Those long tails keep drawing extensions: late-1990s legislation famously stretched existing terms by two decades, keeping early cartoon characters out of the public domain (the first Mickey Mouse film entered it only in 2024). Critics argue that most of the value of very long copyright accrues to heirs and corporations rather than creative incentives, since an author writing in the 2020s is unlikely to be moved by royalties in the 2120s. Extensions raise the private return a little while withholding the work from free public use for a long time - a trade many economists doubt.

Policy two, part A: paying directly

The government can simply fund research itself: grants to university scientists, public laboratories, mission agencies for space, health, agriculture. This plug fits the biggest leak - basic research whose benefits are so diffuse, so far off, or so foundational that no patent could capture them (who patents the laws of physics?). Public money deliberately does NOT capture returns: results flow to everyone, which is the point. The cost is twofold: taxpayers fund it, and grant allocation is political - the scientific question entangles with the geography of swing districts.

Who funds what: the mix

In most rich economies, industry funds the majority of research and development, government a substantial minority, and universities and nonprofits the rest. In the US the federal share fell from about two-thirds in the 1960s to about 18 percent in 2022, when business funded about three-quarters (NSF data). The sorting is roughly rational: industry funds the applied research it can capture; government concentrates where firms are not active - basic science, defence, public health. The pattern is the externality theory drawn in institutional form: the leakier the benefit, the more public the funding must be.

Policy two, part B: the tax credit

Or the government can simply lower the price of research. An R&D tax credit reduces a firm's tax bill in proportion to its qualifying research spending, leaving the choice of what to research entirely decentralised. Most studies of the US research credit conclude that every dollar of tax given up draws out at least a dollar of extra research. The credit's advantage over grants is its blindness: no committee, no politics of project selection. Its weakness is the same blindness: firms relabel ordinary spending as research, so definitions, audits and rules nibble at the design forever.

Policy three: cooperative ventures

The third plug stitches the system together: government-supported research partnerships between universities and companies. Universities generate the basic knowledge firms cannot capture; firms hold the development machinery universities lack; a technology-transfer office in between licenses discoveries and spins out startups. Cooperative ventures also solve a sharing problem: when several firms face a common pre-competitive problem - a new battery chemistry, a semiconductor process - joint research splits the fixed cost and lets the results spread through the industry. The hard design question: who owns the results.

Worked example: matching the tool to the idea

Four research ideas arrive at a ministry; match each to its best instrument. 1. A new battery chemistry with clear commercial uses: industry funds most (capture is high); a patent protects it; the ministry stays out. 2. The mathematics of a new class of error-correcting codes - foundational, decades from any product: basic research, unfundable privately - a public grant fits. 3. Several food firms need a faster cold-chain test, none wanting to fund it alone: a cooperative venture splits the cost and spreads the result. 4. An orphan drug for a rare disease - real social value, tiny market: public funding plus extended exclusivity or prizes. The method: how capturable the benefit, how basic the knowledge, how shared the problem.

Worked example: the patent-term trade-off

A drug costs $500 million to develop and will earn $40 million a year of monopoly profit once patented (an undiscounted illustration). 1. With a 10-year term: expected profit \(10 \times 40 = 400\) million - below cost - so the drug is never developed. 2. With a 20-year term: \(20 \times 40 = 800\) million - comfortably above cost - so development proceeds. 3. The other ledger: patients pay monopoly prices for twenty years instead of ten, with deadweight loss in every protected year. 4. The design problem: long enough to make development rational, short enough to return the discovery quickly. Assumptions: profits constant, no discounting.

Common mistakes to avoid

First, treating IP rights as the opposite of the externality: they are a deliberate partial fence, judged by how much innovation they induce against how much monopoly they cost. Second, assuming public research is automatically political: it is susceptible, like every allocation, and still funds the most leaky, valuable research. Third, reading tax credits as corporate gifts: foregone revenue buys research at least dollar-for-dollar, and decentralisation is the feature. Fourth, confusing copyright with patents: terms, subjects and debates differ. Fifth, expecting one instrument to fit every idea: capturability, basicness and sharedness name the tool.

Key takeaways

Three plugs for one leak. Intellectual property: patents (roughly two decades) and copyright (life plus decades) fence part of the spillover in - imperfectly (inventors capture a third to a half of the value), with the balance-scale patent term (the $500m drug needs 20 years, not 10, to proceed). Direct public funding fits the leakiest research - basic science - at the price of political grant allocation; the funding mix (US federal share about two-thirds in the 1960s, 18 percent in 2022) mirrors capturability. R&D tax credits decentralise project choice while inducing at least a dollar of research per foregone dollar. Cooperative ventures stitch universities to firms and solve shared problems. The matching test: how capturable, how basic, how shared - the answers name the tool.

Flashcards

Flashcard

Three-plug toolbox

IP rights, public R&D funding, cooperative ventures

Every innovation system mixes them.

Test your understanding

Question 1 of 5

Why do patents give a pharmaceutical firm an incentive that pure competition would not?

Question 2 of 5

In the patent-term example, a $500m drug earning $40m a year is not developed under a 10-year term. What does this illustrate?

Question 3 of 5

Why does government directly fund basic research rather than leaving it to firms?

Question 4 of 5

What is the main advantage of R&D tax credits over direct government grants?

Question 5 of 5

Why are cooperative research ventures suited to 'pre-competitive' problems like a shared cold-chain test?

← Previous topicNext topic →