The transformation unleashed by increased funding for science during the 20th century is nothing short of remarkable. In the early 1900s research was a cottage industry mostly funded by private firms and charities. Thomas Edison electrified the world from his industrial lab at Menlo Park. Advances in science during the second world war led governments and companies to scale things up. By the mid-1960s America’s federal government was spending 0.6% of GDP on research funding and the share of overall investment in research and development rose to nearly 3%.

That dynamism is fizzling out. A growing body of work shows that even as the world spends more on research, the benefit-cost ratio has fallen. One explanation for this is that the way science is funded is out of date. Researchers must now contend with amounts of bureaucracy. The rate at which grant applications are accepted has fallen, meaning more of them must be made. Two-fifths of a top scientist’s time is spent on things other than research, such as looking for money. The current system is also monolithic. Western scientific systems are dominated by handouts of project grants and peer review. Most money flows to universities, and the academic career ladder is such that researchers face incentives to pursue incremental advances, in order to boost citations and gain tenure, rather than breakthrough work. It is time for another shake-up.

A growing number of scientists, policymakers and philanthropists hopes to revamp science funding. In 2022 America’s CHIPS Act reformed the National Science Foundation (NSF) to focus it more on technology. America’s famed Advanced Research Projects Agency (today called DARPA) has inspired copycats in Britain and Germany. Tech billionaires’ plans to fund pet projects come thick and fast. On November 1st Eric Schmidt, a former boss of Google, announced he was funding a Moonshot to build an “artificial-intelligence scientist” to speed up biology.

What’s the best ways to fund science? The first step is to try new things. More money could fund promising people rather than specific projects, encouraging researchers to take risks. Funders could move faster and bypass peer review entirely, for example by using lotteries. More important still is to find ways to measure what is working and what is not, and then adapt accordingly. Governments might consider appointing “meta-scientists” or “chief economists” to do the number-crunching across their various scientific agencies.

None of this will be easy. More cash for DARPA-like bodies means less for other approaches. Scientific funders say they want to experiment, but they also face pressure to support research that can be easily explained to keep politicians happy. In some cases more money may be the only solution. Still, the economic returns to research are so large that fixing the system is well worth the effort.

The word “revamp” (Line 2, Para. 3) is closest in meaning to__________.