Who Should Pay for Risky Science?

1For most of the past century, the arrangement behind American research looked simple. Governments paid for basic science, universities and national laboratories carried it out, and companies picked up whatever looked profitable. The arrangement produced extraordinary results, and for that reason it was rarely questioned. Lately, however, a small but vocal group of researchers who study how discovery itself works has begun to ask whether the machine is still fit for purpose.

2Their first complaint concerns how money is handed out. Public funding for basic research tends to pass through a limited number of review panels, and panels, being made up of experts, tend to reward work that experts already find plausible. Proposals that depart sharply from the consensus are, by this account, the ones most likely to be turned down. The biologist Katalin Karikó, whose work on modified messenger RNA later underpinned vaccines against covid-19, is the example critics reach for most often: for years her grant applications were rejected as too speculative. Few researchers, they point out, would have persisted as she did.

3The second complaint concerns time. Scientists who depend on grants spend a large part of their working week writing proposals, reports and renewals rather than doing research. Universities, organised into departments that rarely talk to one another, are also said to be awkward homes for problems that cut across disciplines.

4None of this means private money is a new idea. Some of the most celebrated inventions of the twentieth century, the transistor among them, came out of corporate laboratories rather than university campuses. Yet those laboratories gradually shrank, and the reason was economic rather than scientific. The economist William Nordhaus estimated that firms capture only a small fraction of the value their innovations create; most of it spills over to competitors and consumers. Faced with that arithmetic, many companies stopped running large research laboratories of their own and began buying promising start-ups instead.

5What is new is the shape of the latest private ventures. Some are non-profits that give a team a fixed sum and a few years to build one specific tool, a scope too large for a single academic lab and too uncertain for investors seeking a quick return. Others hand a small number of researchers wide freedom to change course when an approach fails, with no obligation to publish on schedule. Their backers argue that this flexibility, rather than the size of the cheque, is what conventional funding lacks.

6Enthusiasm should be tempered, though. Several such institutes are young, and at least one well-funded venture has already closed within two years of its launch. Many of the newer bodies also expect their work to lead, eventually, to something that can be sold, and much valuable inquiry, from pure mathematics to the history of ideas, will never meet that test. Those questions will remain the business of universities and public funders.

7The more modest conclusion, then, is that the debate is not really about public against private. It is about whether a research system built around a few gatekeepers can afford to make so few bets on ideas that look wrong at first. Private institutes may turn out to be less a replacement for the old machine than a test of which parts of it were holding discovery back.