A draft report by the Productivity Commission reiterates what the federal government has repeatedly said – despite a 14 per cent increase in spending per student over the past decade our performance in national and international assessments has barely improved.

In some cases, it has slipped.

Source: Productivity Commission 

While a lot of data is collected on student and school performance, there is no proper way of evaluating what works in Australian classrooms.

The federal government asked the commission to investigate whether Australia's education policies and practices were evidence-based and how we might create a more rigorous "national education evidence base" to boost student performance.

The commission's draft report, released on Tuesday, found that competition between schools, performance benchmarking and monitoring student outcomes through tests like the NAPLAN were "insufficient to achieve gains in education outcomes".

It said these practices needed to be accompanied by evidence highlighting the most effective policies, programs and teaching methods.

The report also raised concerns about the quality of education research in Australia, and said more than 40 per cent of education research units recently evaluated were rated as below world standard.

It urged high-quality research, particularly randomised controlled trials, which would make it easier to decide whether to discontinue ineffective programs,

"This means education expenditure can be used more effectively and efficiently," it said.

The draft report comes as NAPLAN results continue to flatline, with federal Education Minister Simon Birmingham calling for a focus on "evidence-based measures that will get results for our students".

And Australia's performance in the PISA – an international test that pits 15-year-old students against their overseas counterparts – continues to slip.

Our PISA maths literacy and reading performance dropped by 29 and 16 points between 2000 and 2013. Australia ranks 14th behind Poland, Germany and Vietnam.

The findings are likely to be seized on by the Turnbull government, which has said many times that simply throwing more money at schools will not improve student results.

But critics say this is a convenient argument being used to justify the federal government's decision to abandon the final two years of the Gonski school funding agreement.

Public education advocate Trevor Cobbold, a former Productivity Commission economist, recently wrote a paper accusing the government of using highly selective data to "misrepresent the real state of school funding". He said that when adjusted for inflation, funding per public school student fell by 1.9 per cent between 2009 and 2013 but increased for Catholic and Independent schools by 8 per cent. The figures used in the commission's report refer to funding for all schools.

Productivity Commissioner Jonathan Coppel said the report did not downplay the importance of funding.

"Resources are important," he said. "But the way they are used and how to get the biggest bang for your buck depends on having strong research processes."

He said it was time to find out what good schools were doing, and replicate their methods across other schools.

The commission has recommended that Australian, state and territory governments create a new Education Agreement that defines the aims of commissioning and applying research that evaluates what works best. It also recommended changes to privacy laws to make it easier to access, and sharing de-identified data when evaluating education programs and practices.

An Australian Education Union analysis released on Tuesday found that Australian schools would be $5.3 billion worse off from 2016-17 to 2019-20 under the Turnbull government's funding model, compared with the full six years of Gonski funding.