The productivity of scientists has been stalling, risking economic growth with it. This briefing explores ways in which government policy can address two great challenges facing academic science: falling research productivity and the trustworthiness of its findings.
Academic science faces two major issues:
- Research productivity appears to be declining. Scientific progress appears to be fairly constant, but maintaining this rate of progress is becoming increasingly costly. If the trend continues, we may face a choice between devoting increasingly large proportions of our national income to maintain technological progress, or accepting slower technological progress and a corresponding decline in economic growth.
- There are issues with the replicability, reproducibility, and robustness of large parts of the academic scientific literature. The same incentives push scientists towards pursuing more incremental and less potentially transformative research agendas that encourage bias, hype, fraud and negligence.
- Create shared research infrastructure, such as shared repositories for data, code and papers.
- Enforce UKRI’s Open Research policy, publishing data on the accessibility of publicly-funded research and setting a deadline for full compliance.
- Diversify UKRI funding mechanisms, exploring innovative approaches such as:
- Funding people not projects, with fellowships to promising young researchers
- Giving reviewers ‘golden tickets’ to fund radical ideas
- Run trials of funding mechanisms, such as assigning funding by lottery
- Establish the ‘National Institute for Scientific Replicability’, which would sponsor replication projects and develop tools to predict replicability
- Establish the ‘Atlas Institute’: an organisation which maps scientific progress and encourages scientific productivity and interdisciplinary research.