I’m at the new Education Policy Institute offices at 9am on a sweltering Monday to meet its executive chairman, the former Liberal Democrat schools minister David Laws, and Natalie Perera, an ex-civil servant turned executive director. I’m envisioning some kind of breakfast banter over croissants, a la Richard and Judy, or maybe more Fern and Phillip.

As it is, the relationship between the two – who have worked together for coming on six years, first within government but now lobbying from without – has less repartee and more gentle respect. Perera always lets her boss finish what he’s saying, even when you can see she’s itching to speak, though more surprisingly perhaps, Laws never interrupts his executive director.

The second part of Perera’s title is “head of research”, which seems an unusual role for someone who is not university-trained, I venture.

“Natalie is a great example of the fact you don’t have to go to university to really excel in things and reach the top,” Laws counters, with disarmingly contagious calm confidence. “I always assumed when we were in the department that she had about three PhDs, she used to blind all of us so much with her knowledge and data. It just shows what can be achieved if you’ve got the drive and motivation and talent.”

The appointment also reveals EPI’s priorities. Located by Victoria station, just two tube stops from Westminster, the think-tank is unabashed about its aim to directly affect education policy, and it has recruited to senior posts a further four or five “really talented, motivated bright people” from the DfE.

The former minister denies that he has followed a deliberate strategy to draw from the civil servant talent pool. It’s simply, he suggests, “the way it worked out” – perhaps a little disingenuously, since he goes on to admit that it’s very convenient they should have intimate working knowledge of the national pupil database.

For her part, Perera is keen to stress the importance of growing their own talent, too – through a graduate internship programme, among other routes.

The EPI positions itself as a bridge between university academics and the government, using strategic partnerships “to elevate academic research in a way that doesn’t usually have that kind of profile,” she says.

I almost never received the sort of quantitative research reports that might be helpful to me as a minister

This seems a natural progression for Laws, who has often discussed his frustration at the degree to which policy preceded evidence during his time as a minister in the Coalition era. He echoes this now. “I almost never, in my red box as a minister, received the sort of quantitative research reports looking at issues in education and producing conclusions that might be helpful to me as a minister.”

Michael Gove, his superior at the Department for Education, had a reputation for a cavalier attitude to evidence. So is Laws – whose 2016 book Coalition revealed just how charmed he was by his boss – willing to throw him under the bus on this particular question?

“I think Michael had both a very strong interest in data analysis and evidence,” he responds with diplomacy, “but also unusually, that was mixed in with some very strong personal opinions, based upon his experience.”

The problematic tendency of politicians to “leap from personal experience to drawing conclusions for the system as a whole” is affecting Number 10’s current support for grammar school expansion, Laws believes.

“If you’re Theresa May or some of the other Tories who have been to grammar schools, and you did well, and got to the top of your profession, or became prime minister or something, it’s very tempting for them to think ‘well, obviously that’s a really great way of delivering social mobility’.

“There’s a risk of translating what appears to work for one individual or a small number of individuals and concluding that it is the way to design the entire system. That’s clearly a fallacy.”

Perera is keen to add her take.

“What I think proponents of grammar schools don’t quite understand is, even those disadvantaged children who do get in and do see a third of a grade higher in their results are likely to be a unique set.

“We don’t know, for example, whether their parents are graduates and have just fallen on hard times, or whether they are from immigrant families who in other countries might have been lawyers or whatever.

“So I think we need to be careful in over-egging what this very tiny group of disadvantaged kids in grammar schools truly attain.”

In Gove, a passion for evidence coexisted with very personal biases. Laws recalls how he would invite people like Andreas Schleicher of the OECD to brief his education ministers on trends in international education, and the implications for policy. However, he also had some “deep-seated beliefs about things like the value of autonomy in improving educational standards in schools, and whether a 19th-century English novel should be compulsory or not”.

“It’s quite unusual in my experience to have those two things coexisting in one person, but then,” he adds, echoing the somewhat obsequious tone of his book: “Michael is quite an unusual and exceptional person.”

David Laws and Natalie Perera at the EPI offices

Perera’s analysis from 12 years spent inside the Department for Education, is that there is often a gap between the talented people “tucked away doing lots of great research and analysis” and the policy advisers speaking to the ministers. While her tone is neutral, conciliatory even, her words suggest a somewhat chaotic environment.

“It’s a big department, so the research could go upwards, you know, to the minister, but there were lots of intermediaries where it could go, or indeed nowhere at all,” she says.

Laws is gently scathing about the rigour of the UK’s use of evidence in political decision-making, even in comparison with much poorer countries. He contrasts Liberia, which ran a $1 million randomised controlled trial at a small number of schools before introducing its own academy-style system, to the ad-hoc way academisation was introduced in England – which of course, didn’t have international donors to account to.

“The great irony for me, working with an organisation that looks at education policy in the developing world, is that quite often, the degree of rigour, evaluation, research in very poor countries is greater than in developed countries like ours, often because the interventions are funded by big foundations that insist that there is evaluation and research alongside them,” he says.

Research in very poor countries is greater than in developed countries like ours

This example nicely illustrates EPI’s mission, which is to be “for education what the IFS is for the fiscal policy space,” explains Laws, “a highly quantitative, rigorous, independent, data-led organisation, but one that does want to have an impact on policy, and ultimately on outcomes”.

Perera concurs: “one of the things we want to avoid is plucking solutions out of the air that aren’t evidence-based.”

Some may call this dream for evidence-led policy naive. In fact, the policy U-turn for which Laws’ Lib Dems will be forever excoriated by an entire generation of millennials – that they allowed their Tory coalition partners to massively raise university tuition fees in exchange for investment in early-years education – was a classic example of the dangers of imposing an evidence-based policy (ie that more money for the early years would do more to close the disadvantage gap than ditching tuition fees) without first convincing your grass roots.

No longer an MP, if Laws ever returned to the education department – which he stresses is “immensely unlikely” – the early years would be an integral part of his plan to improve the quality of policymaking.

“I wouldn’t have ministers responsible for early years, schools and post-16. I really don’t think that leads to joined-up policymaking,” he declares. He would instead allocate ministerial responsibility by theme, so there would be one minister responsible for “improving quality of learning” throughout all the education phases, for example, and another for “capital and revenue funding”.

“I think the way the DfE was and is, to some extent, broken up between age phases, is actually quite corrosive of good education policy,” he muses. “If the schools minister were responsible for early-years quality, there would be a much more joined-up process.”