International tests reveal the performance of our schoolchildren is plummeting despite years of education reform designed to create brighter futures for the next generation. Simon Collins asks the experts where it all went wrong.

Emeritus Professor Warwick Elley worries that New Zealand's education system is failing an entire generation.

"I worry that it's a dumbing down of a whole population of students," he says.

When Elley chaired the international steering committee for one of the first world literacy surveys, in 1990, Kiwi students came fourth.

A decade later, when the Programme for International Students Assessment (Pisa) started testing 15-year-olds, NZ students came second only to Finland in reading, third in maths, and sixth-equal in science.


But it has been downhill ever since. In six three-yearly Pisa surveys, the most recent (2015) reported last December, each group of NZ students has scored lower than the group that went before them in both reading and maths.

Over Pisa's 15-year history New Zealand's average score for maths has dropped by more than any other country (down 42 points), closely followed by Australia (down 39 points).

Our average for reading has dropped by 20 points, a steeper fall than in all except three countries (Britain, Australia and Iceland).

Even in science, where we have had ups as well as downs, our average is down 15 points since 2000, although eight other countries including Australia declined more.

These scores are based on tests in which about half the questions are repeated in every survey (and kept a closely guarded secret) so that each new group of 15-year-olds can be compared with those who came before.

Professor John Hattie, formerly of the University of Auckland and now director of the Melbourne Education Research Institute, says both New Zealand and Australia have been too complacent.

"We have sat back on our laurels," he says.

New Zealand is still in the top half of the rich-nations' club, the Organisation for Economic Co-operation and Development (OECD), on all three subjects: sixth out of 35 nations in science, eighth in reading, 16th in maths. It's the trend, not the level, that is the main concern.


Marks in the Pisa tests were scaled to an OECD average of 500 when they started in 2000. Since then we have dropped from 528 to 513 in science, from 529 to 509 in reading, and from 537 to 495 in maths - staying above the OECD maths average only because it has slipped from 500 to 491.

Four other points may help to understand the trends.

First, Asian countries have consistently been the top scorers in maths and science and, since 2012, even in reading despite the disadvantage of more complex scripts. In 2015, Singapore became the first nation to top all three subjects.

This is not just because Singapore's 15-year-olds study for an average 22 hours a week doing homework and private tuition outside school, compared with 17 hours in New Zealand. Japan also beats us in all three subjects, but their 15-year-olds study outside school for only 14 hours a week.

Second, the three nations that have fallen furthest since Pisa began are all Anglo-Saxon: in order, Britain, Australia and New Zealand. The almost identical tracks of Australia and NZ suggest that there may be common factors driving us both down.

The biggest gains, among those who have been in Pisa since it started, have all been in Europe: Luxembourg, Portugal, Poland and Germany.

Third, New Zealand's graphs were fairly flat or (in science) rising between 2003 and 2009, then dropped heavily across all three subjects in 2012. The Education Ministry believes they may now have levelled out, with much smaller declines in 2015.

"My reading of it is that the decline has halted," says the ministry's deputy secretary Dr Craig Jones.

And fourth, New Zealand has consistently had some of the widest gaps between top and bottom students. The gap between our top and bottom 10ths of students in 2015 was the OECD's second-widest in science, fourth-widest in reading and 15th-widest in maths.

Our gap in science has grown even wider over time.

But our gaps in reading and maths have narrowed because our top students have slipped more than our bottom students. Our bottom 10th actually scored higher in maths in 2015 than in 2012, while our top 10th's scores dropped.

Our gaps between low and high socio-economic groups, and between Maori and Pacific students and the national average, have narrowed across all three subjects.

One possible factor in our decline is that our two lowest-performing ethnic groups have grown - Maori from 18 per cent of 15-year-old students in 2000 to 24 per cent in 2015, and Pasifika from 7 per cent to 10 per cent.

Although our ethnic gaps have narrowed, they are still huge. For example in reading, where our national average is 509, Maori (465) and Pasifika students (450) score below all OECD nations except Slovakia, Chile, Mexico and Turkey. Our wide gaps are dragging us down.

But the fact that Australia's scores have declined almost identically to ours suggests that the biggest drivers are broader. The main candidates are education policy changes in the Anglo-Saxon nations.

In our case, the two biggest changes since Pisa started have been the introduction of the National Certificate of Educational Achievement (NCEA) in secondary schools from 2002, and of National Standards in primary schools from 2010.

Elley points to declining Pisa scores in all five OECD nations that have standards-based assessments with "high stakes", with results published in league tables: New Zealand, Australia, Britain, the United States and Sweden.

"The majority of OECD countries don't allow league tables," he says.

Both NCEA and national standards were born with good intentions - to close our shamefully wide gaps by focusing schools on ensuring that every child achieves the standards.

But Elley believes they have actually dragged most students down. They have:

Focused teachers narrowly on teaching what is tested (only reading, writing and maths in the case of national standards), so students are flummoxed when faced with Pisa's tests of broader knowledge and skills;

Encouraged schools to steer weaker students into easier NCEA subjects that they can pass, such as statistics instead of algebra, so they can't cope with Pisa's harder questions;

Broken up subjects into small chunks for NCEA credits, rather than helping students achieve the deep understanding that comes from seeing the big picture;

Allowed our top students to relax as soon as they reach the standards or gain 80 NCEA credits, instead of stretching them to achieve their full potential;

Intensified competition between schools, so the best schools attract the best teachers and students at the cost of declining quality in other schools.

The business-funded NZ Initiative think-tank points to another set of changes that have combined to weaken teachers' effectiveness, especially in maths where a 2010 study found that a third of new primary teachers could not add 7/18 and 1/9.

Teacher training, which was confined to six teachers' colleges until the 1990s, was deregulated and by 2005 teaching courses were offered at 27 institutions. The six original colleges were taken over by universities.

Lisa Rodgers, a former ministry deputy secretary, who now heads the Australian Institute for Teaching and School Leadership, says the universities have expanded their teaching courses because they are "a cheap course to run".

"Bums on seats equals dollars, and they are not as costly as medical or engineering students," she says.

Massey University Institute of Education head Professor John O'Neill says graduate teachers have become casualised with only 15 to 30 per cent going into permanent teaching jobs, and too often with academic rather than practical training.

"Over the years, teacher education moving to the universities means that very few teacher educators have any recent experience of being in a classroom," he says.

Although the NZ Initiative found that teachers' salaries have more or less kept their historical ratios to other professions since 2000, it says teachers are often lured into other sectors once they reach the top of the teaching pay scale - especially if they have skills in maths and technology, which are in high demand elsewhere.

Finally, the Initiative and Hattie blame a specific teaching change called the Numeracy Project for our uniquely sharp decline in Pisa maths scores. The project, launched in 2000, trained primary teachers to give students "a range of different ways to solve problems" and is said to have reduced emphasis on basic knowledge such as the times tables.

"Teachers loved it because they didn't have to know the maths, they only had to know the strategies," says Hattie.

"It led to 10 years of the greatest decline in maths. People have woken up to the fact that that was one of our biggest failures."

Surprisingly, Australia and New Zealand are the eighth and ninth most equal OECD countries on the Pisa index of socio-economic status, based on parents' education and occupations and their books and other possessions in the home.

Only South Korea, Japan, three Scandinavian nations, the Netherlands and Estonia are more equal.

So our wide gaps in 15-year-old achievement are not because our students start less equal than others; they are because our schools are less effective at closing the gaps that students start with.

Equally surprisingly, despite our intensely competitive schools, our achievement gaps between schools are well below the OECD average, mainly because many OECD countries stream students into separate academic and technical schools.

In contrast, we have the OECD's widest achievement gaps within schools.

Hattie believes this is because we stream students within schools. Ninety per cent of our schools in the Pisa survey stream based on ability levels for at least some subjects. Australian schools are close behind us at 88 per cent. The OECD average is 46 per cent.

"The problem with streaming is that teachers have expectations of what kids can do," he says.

"Imagine a kid that develops later. They can never catch up because they haven't got access [to higher-level subject content]."

Another key element in our wide gaps is that our disadvantaged students tend to be marked out by their ethnicity. Maori secondary school students told Waikato University researchers Russell Bishop and Mere Berryman in 2001 that many of their teachers had low expectations for their achievement and behaviour.

Education Minister Hekia Parata believes we are on track to reversing our declining Pisa scores and narrowing our achievement gaps. Far from blaming NCEA, she celebrates extraordinary gains since she set a target of 85 per cent of 18-year-olds having NCEA level 2 by this year.

Preliminary 2016 results show the target has been achieved (85.2 per cent, up 8.9 points since 2011). Ethnic gaps have narrowed dramatically, with Maori up twice as fast (up 17.6 points) to 74.7 per cent, and Pasifika up 14.8 points to 80.3 per cent.

In Rodgers' words: "The incredible change in achievement over the last few years for Maori and Pasifika is just powerful."

Although Labour's shadow education minister, Chris Hipkins, would scrap national standards, no one proposes abolishing NCEA.

Hipkins would also keep one of Parata's proudest innovations - communities of learning, which bring preschools, primary and secondary schools together to share expertise and to plan "personalised learning pathways" for children from age 0 to 18. Sixty per cent of schools are now in them and Parata is confident all schools will join them in the next two years.

"Schools that participate in a community of learning will have access to secondary schools with technical knowledge in the subject [such as maths] and will be able to be used as a resource," she says.

Parata and Hipkins agree with Hattie that streaming students has been harmful, and support more personalised learning.

Berryman has recently drawn from hui involving 158 successful Maori secondary students to note that, although the students still face "negative stereotypes about being Maori", they now have support from "a school environment where their own culture and values were explicitly celebrated, modelled and thus made more acceptable".

But we still have a long way to go.

"While these students were successful, they knew they were not the norm," Berryman writes.

"Their peers, their whanau and their friends were not able to overcome the barriers and were not achieving the success that is the expected outcome of schooling."

Stumped by a revolving door

It was a question about a revolving door that stumped our group of Kiwi 15-year-olds.

Our five Year 11 students at Waiuku College who had agreed to try out sample questions from the Pisa tests sailed through the online science questions, which tested their understanding of evolution and their ability to apply a given formula to a computer-simulated case study.

From left: Waiuku College students Nic Ruygrok, Zach Whitley, Latesha Latu, Mikayla Capes and Mauriora Kaihau. Photo / Jason Oxenham
From left: Waiuku College students Nic Ruygrok, Zach Whitley, Latesha Latu, Mikayla Capes and Mauriora Kaihau. Photo / Jason Oxenham

Most also got the right answers to questions testing simple arithmetic.

Then they hit the revolving door.

"A revolving door includes three wings which rotate within a circular-shaped space," the question said.

"The inside diameter of this space is 2m (200cm). The three door wings divide the space into three equal sectors."

The first part of the question was: "What is the size in degrees of the angle formed by two door wings?"

The answer required knowing that there were 360 degrees in a circle, and dividing by 3. Only one student got it right.

The second part said: "The two door openings [inside and outside] are the same size. If these openings are too wide the revolving wings cannot provide a sealed space and air could then flow freely between the entrance and the exit, causing unwanted heat loss or gain. What is the maximum arc length in centimetres that each door opening can have, so that air never flows freely between the entrance and the exit?"

To get the right answer, students had to know the formula for the circumference of a circle (2 x pi x r), where pi is a long number starting with the digits 3.14 and "r" is the radius of the circle (100cm in this case). Only one student got that far.

Then they had to divide by 6 because the maximum size of each door opening was one-sixth of the circumference. Even the one student who knew the formula stumbled at this point.

"We forgot the actual formula," said Mikayla Capes, who had done well until then.

The consensus, as Mauriora Kaihau put it, was: "The science was easier than the maths."

Maths was the weakest Pisa subject for Kiwi 15-year-olds generally, and the revolving door question suggests that what we lack is not the thinking ability but the basic rote-learned knowledge - in this case, the formula.

The students had their own views on why Kiwi scores were declining.

"Maybe they are not taught how they need to be taught," said Latesha Latu.

"Sometimes I feel it can be tough on both ends [of a class] - the in-betweens are getting what they need, but the high end and the low end are not being catered for enough."

Waiuku deputy principal Gowan Ditchburn, who supervised our test, said every school faced the dilemma of coping with students of differing abilities.

At Waiuku they start streaming top students into an extension class in Year 10. Those who were not picked for it, like Nic Ruygrok, said it "kind of made us feel inferior".

Zach Whitley, who was also left out, said: "Say you're talking to your friends [in the extension class] and they have an assignment due, and sometimes they are worth credits. I could do them with my eyes closed, but [pupils in regular classes] don't really get the opportunity."

Ditchburn said streaming also reduced the chances for slower students to learn from the faster ones - a practice that helps the faster students understand what they are learning, as well as pulling up the slower ones.

"Equally, there are other arguments about teachers teaching to the middle and therefore not meeting the needs of the students at the extremes [in non-streamed classes]," he said.

"Personally I can see arguments for and against on both sides of the debate."