In 2013, when my colleagues and I began a three-year study of how evidence is used in the UK Parliament, Brexit wasn’t a thing, Donald Trump wasn’t a politician and ‘fake news’ wasn’t in the news. By the time we finished, it felt like we were living in a different world, with democracy creaking under the burden of misinformation.

The tumult was good for our research, at least. Politicians and policy advisers were increasingly thinking about the role of evidence, and we benefited from unusually unguarded introspection. Some politicians admitted that they use evidence to bolster political positions and score political points. Overall, though, there was a healthy appetite in the UK Parliament for evidence to inform decisions.

It is an appetite that often goes unsatisfied. Our report, released on 30 November, found that academic research just isn’t getting through. Most of the evidence supplied to the legislature comes from government and non-profit civil organizations. Academia needs to step up, and research funders should make sure that it happens.

Why is this important? Democratic decisions need factual foundations. How should we reduce obesity? Do we need more nuclear-power stations? Should we prioritize welfare or tax cuts? Answers require value judgements, but these should be underpinned by evidence.

For the most part, the UK Parliament has good systems for using evidence. Multidisciplinary units of politically impartial staff endeavour to answer all manner of questions for politicians, such us how planning regulations work or whether Wi-Fi radiation is safe. I spent five years directing one such unit, the Parliamentary Office of Science and Technology (POST), that proactively communicates academic research to politicians. Many legislatures globally have similar set-ups.

The problem is that equivalent infrastructure on the academic side is lacking. Since 2007, all UK Research Councils have required funding proposals to include statements describing ‘pathways to impact’. But researchers and reviewers attach little importance to them.

No wonder people working in policy can feel that academic engagement is clumsy and naive. In my two stints working in Parliament (I also advised the House of Commons Science and Technology Select Committee in 2007–10), I have been repeatedly frustrated by the poor quality of academic engagement. It ranges from jargon-laden ‘briefs’ that contain little pertinent information to condescending homilies about what democratically elected politicians should care about.

For research truly to inform policy, it is not enough to hope that the stars will align. The stars need to be wrested into position.

Here’s an example of how things can work. In 2015, the United Kingdom’s legislature became the first in the world to legalize an in vitro fertilization technique that prevents mitochondrial disease by combining components of two women’s eggs. The technology was anticipated and provisions were written into regulations back in 2008. As the technique matured, POST worked in tandem with outside experts to answer politicians’ questions and concerns.

But most policy issues are immeasurably more complex than mitochondrial transfer. More coordination and long-term engagement are typically required.

To make sure that capacity exists, research funders should support policy-relevant work only when scientists have given serious thought to policy impact. Funders from both the government and charitable sectors have told me that they’d like to fund innovative policy activities, but the proposals from academia just aren’t there. They would be, if funders demanded them.

This proposal will raise eyebrows. First, how does one distinguish between policy-relevant research and other kinds of work? Second, how would one evaluate whether a policy-impact plan is good or poor? I am confident that these problems can be solved by a community that identifies tiny differences in research excellence.

The answer to the first question could be a combination of self-identification by researchers, perhaps to qualify for extra funded activities, and peer review. The answer to the second could be partly criterion-based. For example, good proposals will be written in conversation with policymakers, who will have helped to set research questions and perhaps even the research plans. Fundable proposals will describe when and how policymakers will be involved in the work, either to provide ongoing engagement to the policy problem, or to be kept at arm’s length to avoid biasing the research process. Good proposals will also contain concrete outputs for policymakers, such as briefs, reports and meetings. Finally, systems should be put in place so that engagement with policymakers can continue for years after the research project is completed.

Funder mandates will force universities to take policy impact seriously. Institutions’ current activities in this area are disparate, poorly funded and often ineffective. Instead, research-intensive universities should set up dedicated policy-impact units. These should be staffed by professionals who are adept at navigating academia and policy. They should work across disciplines and universities, and provide a mix of proactive and responsive advice to connect research findings to policy needs.

The academic community has a duty to ensure that research evidence is brought to bear on legislation to keep our democracies healthy.