Performance of New Zealand students in math, science, and reading falls dramatically in last two decades

December 27, 2021 • 11:00 am

Why should we care about the performance of New Zealand’s primary- and secondary-school students, and what’s happening with it over time? For me, it’s the science that’s important, but science, reading, and math show the same trend over the last fifteen years. Despite a rise in spending per pupil over the last 25 years, performance in these three areas in New Zealand has declined, both absolutely and in comparison to the countries like England, Australia, the U.S., Canada, and Singapore—countries regarded as educational competitors with (and comparable to) New Zealand.

Why does this matter? For science, at least, as I’ve written repeatedly, there’s a big-time initiative in New Zealand to have mātauranga Māori, or Māori “ways of knowing”, taught as coequal to modern science in the science classroom. This initiative, propelled by the desire to buttress an oppressed minority (the native Māori), has good intentions behind it—to get more Māori interested in science—but is a practical disaster. That’s because mātauranga Māori is not only “traditional practical knowledge” (e.g., navigation, growing crops, catching fish), which can be considered “science construed broadly” (but do you need to teach this in science class?) but, worse, a mixture of legends, myths, morality, and philosophy, some of which is palpably false. Much of it is simply not science as the modern world knows it.

Mātauranga Māori involves, for instance, straight-up creationism of life and the cosmos. You can imagine if students are taught that falsehood alongside biological evolution in class. The teacher, of course, wouldn’t be allowed to say that the mātauranga Māori version is false, for that is disparaging the indigenous people.

What will happen if mātauranga Māori is taught as coequal to and as good as modern science is that both Māori and non-Māori students will get confused about science, and performance on international tests will decline. In fact, it’s been declining for some tme, so now is not the time to drag any traditional “ways of knowing” into the science class.

Now mātauranga Māori should be taught in some venue, but the science classroom is not the place—especially if it gets equal time with modern science. No, it should be taught in sociology, anthropology, and history class, and it should be taught for the same reason that we teach (or at least should teach) about the history of Native Americans in North American schools. It’s part of the country’s heritage and history.

Here I’ll document briefly the absolute and relative decline of student performance in reading, math, and science since the mid-1990s in New Zealand (henceforth NZ). The article below, which I’ve been referred to repeatedly when inquiring, has the data for these three areas; it’s from the New Zealand Initiative, which characterizes itself as “New Zealand’s leading think tank”:

This is what The New Zealand Initiative is all about. We are the organisation to sketch pathways towards a better future. Our mission is to help create a competitive, open and dynamic economy and a free, prosperous, fair, and cohesive society.

As New Zealand’s leading think tank, we work closely with our members, policymakers across the political spectrum, the wider business community, the media, academics and the general public.

Our researchers conduct independent research on a wide range of policy issues. From education to economic policy, from poverty to housing, and from local government to immigration, we are injecting new ideas into New Zealand’s political debates.

We are strictly non-partisan in our work and welcome an open exchange of views and ideas. The results of our research are made available to the public, free of charge, on our website.

Click on the screenshot to read the article, and you can also download a pdf.  Both of the authors work for the Initiative.

I’ll just show a bunch of graphs. First, the conclusion and the three tests used:

The analysis shows that both primary and secondary students’ performance has declined over recent decades. As a result, our international rankings in reading, maths and science have slipped, in some cases markedly. At the same time, New Zealand’s per-pupil education spending on primary and secondary students has increased substantially, both in absolute and relative terms.

It appears our additional investment has not borne fruit, and we should not necessarily expect it will in the future. Indeed, OECD analysis suggests there is virtually no relationship between per-pupil spending and achievement beyond a certain level of spending, a level New Zealand has surpassed. Educational performance.

. . . The three international education surveys used in this report to study and discern patterns in New Zealand’s educational performance are the Progress in International Reading Literacy Study (PIRLS); the Trends in Mathematics and Science Study (TIMSS); and the Programme for International Student Assessment (PISA).

There’s also the NCEA, or National Certificate of Educational Achievement, an internal certification which I presume is equivalent to graduating from American high school with a given ranking. The three ranks are, in descending achievement, 1, 2, and 3.  As the government qualification body states, “NCEA Level 2 has become an important and well-regarded qualification and is often a necessary requirement for the entry-level of jobs.” That’s why they use level 2 in the last figure below.

The National Certificate of Educational Achievement (NCEA) is the main national qualification for secondary school students in New Zealand.

NCEA is recognised by employers, and used for selection by universities and polytechnics, both in New Zealand and overseas.

On to the time graph of performance relative to other countries (Singapore is always tops):

Reading literacy. Note that New Zealand has dropped over 15 years, and is now lowest of all six countries compared:

Math and science achievement in New Zealand itself over 24 years. Both rose compared to the 1995 time point, but then have fallen (science more than math) between 2003 and 2019. Remember, this is not a comparison with other countries, but still uses an internationally standardized test:

This is the most depressing: a fall in year 11 (near the end of school) achievement in reading, science and math on an internationally standardized test:

This has happened despite a fairly hefty rise in per-capita spending per pupal in NZ. Throwing money at schools doesn’t guarantee higher achievement.


Finally, the most depressing graph at all, showing that while NZ has dropped in reading, science, and math on international tests, the percentage of students leaving school with their NCEA certificates at level 2 or better (1) has grown from nearly 60% to 80% in the last 17 years. You’d conclude from the red line that student performance has increased, but what’s probably happened is that the NCEA standards have decreased at the same time that NZ students are doing worse in its constituent parts when assessed using international tests:


There are a lot more data in the report, with comparisons of many other countries besides the six above.

Why has this happened? Well, the summary piece from the Initiative below (click on screenshot) suggests that the rise of student-centered educational design—that is, teaching what the students demand to be taught (or not taught)—had led to the decline. (Click on screenshot). 

From Lipson:

Over the past few decades, the national curriculum and assessment have turned the school system into an experiment in child-centred orthodoxy.

The philosophy has changed everything from what is taught to the teacher’s role in the classroom. It has transformed the purpose of school.

By appealing to the inarguable idea that children should be at the centre of decisions about their learning, children-centred orthodoxy has undermined subject knowledge. It has told teachers they are at their most professional when they let their students lead.

Consequently, educational standards have plummeted. Despite a 32% real rise in per-pupil spending since 2001, students have gone from world-leading to decidedly average.

In reading, maths and science students now perform far worse than the previous generation just eighteen years ago. In all three subjects, 15-year-olds have lost the equivalent of between three and six terms’ worth of schooling. Far fewer pupils today perform at the highest levels. Far more lack the most basic proficiency.

Worse, in the latest round of OECD testing, New Zealand recorded the strongest relationship between socioeconomic background and educational performance of all its comparator English-speaking countries.

Lipson then reproduces the last figure above and says this:

Yet, without these international metrics, there would be no way to see this systemic failure. In fact, so strong is the grip of child-centred orthodoxy that the data from the national assessment, NCEA, shows the opposite.

Now I don’t know enough about NZ schools to tell if Lipson is right. What is important for our purpose is this:

Proficiency in reading, math, and science in NZ has been dropping over two decades.  Teaching mātauranga Māori in science class as a “way of knowing” coequal to modern science is a recipe for disaster, driving science scores even lower in internationally standardized tests.

People of New Zealand: do you want your future citizens to be far less literate in science than they are now?

34 thoughts on “Performance of New Zealand students in math, science, and reading falls dramatically in last two decades

  1. [Grumpy mathematician advisory]. All of the graphs, except for the expenditure per pupil, are plotted with y axes that start far above zero. This exaggerates the differences between the categories and is highly misleading. In the first graph, it might appear that Singapore students achieve double the score of their New Zealand counterparts, whereas the reality is 580 versus 520.

    1. A valid concern for many scientific graphs, but not here. In these types of studies, “zero” has little meaning. It’s the spread around a baseline, in this case 500, which is clearly marked in red. Each data point even has the number explicitly written next to it, so it should be obvious what the deal is. The relevant information are anyway not the numbers, but the relative performance and spread between countries. The real problem, in my view (as a physicist) is that we have no error bars. So it’s difficult to assess how meaningful the trends are. But at least they are fairly monotonic, so we’re not just looking at noise. One serious issue, though, is that countries over the years can learn how to improve in PISA and similar tests. For example, in graph 1, Singapore’s jump from 2001 to 2006 is large. Did they massively change methods on such a short timescale and immediately reap the benefits? I’m doubtful. Education is a bit of a supertanker.

      1. Well, if I may be an even grumpier mathematician, it IS the real magnitude of the change that should concern us, not merely its direction or its uncertainty. The uncertainty or error bars would indeed help us judge the quality of the estimate of this magnitude, but in the end, it’s the magnitude of the change that counts.

        Now it may be that this measure is a poor measure of any meaningful magnitude. That is a common problem in the “softer” sciences. A poor measure may not have any meaningful interpretation, and then we are indeed only left with relative statistical statements (perhaps expressible only relative to the within-country spread).

        1. To interpret the magnitude, you would have to know what 580 vs 500 actually means. That might be possible to some extent by carefully studying how these studies were set up, and how the results were cast into the numbers we see. There is simply no way that the graphs and a caption can do that. So no, the magnitude is not easily useful. That’s why comparison to other countries (and to a lesser extent, the time series of a single country) is all we can readily do. I see no way out of that. In any case, is 580 “good”? Who defines that? What does it mean to “do well”? But if you get 580 and all other countries are below, say, 540, and the average is somewhere around 500, then yes, you can conclude that given the circumstances you are doing well. Never mind my previous observation that countries/school systems can practice for these outcomes. And obviously, tests will always favour rote learning. So all this has to be taken with a grain of salt.

          1. “So no, the magnitude is not easily useful”

            I think you may be misunderstanding my point. As I said, perhaps the magnitude of THIS measure is not easily interpretable, except statistically. But if so, that is a fault of the measure and the study. We really do need to know a magnitude (on some meaningful scale) in order to interpret the result. It may well be that this particular measure is splitting hairs, and this amount of variation does not matter much.

            I agree that it is really hard to find measures with meaningful magnitudes in the soft sciences. But that should be the goal.

            1. Agreed. This is the curse of synthetic scores. Similar to, say, the Gini index, and unlike the GDP per capita where you can have a pretty decent notion what that might imply off the bat.

    2. The scores are not increasing from zero. They are based on a scale center-point of 500 (which was the mean score from 2001 in the case of PIRLS). The difference between 580 and 520 is very large.

      1. NZ is not a woke country for Asians. I write from sunny Auckland, NZ, where last night I went to the movies to see Spielberg’s ‘West Side Story’, since there is no O running riot here at the moment. Readers may be interested for some hard facts about NZ from someone like me, born in Asia but educated since primary school in NZ, and neither White nor Maori. [ Therefore, I do not share the postcolonial guilt of White NZers ]. In NZ, 15+% of the population has Asian ethnicity, and 11% has non-Maori Pacific Island ethnicity. White NZ postcolonial guilt has generously been enforced on the non-White and White recent migrants to NZ, with the results as chronicled in the WEIT blog this year.

        I attended NZ medical school 25+ years ago, in the ‘glory years’ when there was no academic discrimination against Asians. Back then in my med school, 20% of places were reserved for ethnic Maori and Pacific Island candidates, marks for this cohort based on separate test results. [ What is a Maori for university purposes? Someone who is accepted by at least one Maori tribe as theirs, according to their criteria. Google ‘Peter Robinson 3.125% Maori artist.’ Google search should show this University of Auckland painter is satirising his 3.125% Maori ethnicity to take the piss at the definitional status for Maori benefits. He is classified by Auckland Art Gallery as a ‘Maori artist’, despite the satire.]

        Med school entry in NZ follows the UK system, ie straight from high school, plus also after a basic university degree if the direct route fails. In my final year high school exams I obtained what was known as a ‘junior scholarship’ ie total marks in the top 100-150 of all high school students that year, of which around 30-40 applied for med school. This allowed me what was called ‘preferential entry’, where I didn’t need to compete in my freshman year [ so I studied philosophy, , literature, religion, and biology ]. Back then, only around 3% of the NZ population were Asian. As the Asian % of NZ rocketed, more Asians were successful for med school, placing pressure on White entrants. Interviews were introduced at first, and then made more important as a criterion for entry. About 20 years ago, ‘North and South’ monthly magazine ran feature on med school entry which showed photos of that year’s Auckland med school graduates [ more than half Asian ], versus around 1970. The article’s thrust was that the Asians were becoming overly successful.

        Since the interviews had no criteria except what the largely White and Maori interviewers felt best applied to being a medic, the % of Asian NZ med students reduced somewhat from the peaks of ‘marks only’ . In addition, the % of the med school class reserved for Maori and Pacific slowly increased. Now at Otago med school, there is a priority intake criterion : 1st priority the Maori/PI group, then smaller groups for ‘rurally-based students’, a couple of disabled, and one or two students from refugee backgrounds. The remainder are called ‘general category’, where most Asians fight it out with most Whites [ ie the Whites who could not ransack their family tree for at least 3.125% Maori ethnicity ]. In 2019 for Otago med school, the first year university % marks minimum threshold was around 74% for the Maori & PI reserved group, while the general category required a minimum entry of 94% in the same university exams.

        Dr Lipson in the article is a good researcher, I believe originally from the UK. The NZI think tank is a right wing centre roughly equivalent to the US AEI or Hoover Institution. Its ideological slant for education is to tout the benefits of charter schools and private pay-money schools. Therefore its solutions are slanted to align with its ideological goals. However, the problem with anti-Asian discrimination is neither right wing nor left wing.

        The actual rot in the calibre of NZ students is not so much evident in science/maths [ where the top 5% of high school students are simply better than the top 5% in my equivalent year ], but in the arts. Readers may know how liberal arts degrees are touted as ‘degrees in critical thinking’. Sadly in NZ, the people who are most woke in all the negative senses are NZ arts degree graduates. [ From my personal experience of NZ culture for the past 30 years, non-NZ arts graduates have a better purchase on reality. ] Check out the landing page of Auckland City Council culture & events, produced I believe by NZ arts graduates –

        Note paragraph 3 : ‘welcome to tamakai makaurau Auckland an indigenous city of the Pacific’. This year I wrote a letter to the editor of the NZ Herald [ the main national newspaper ] complaining of the lack of critical thinking on this website. I pointed out Auckland’s population is 28% Asian, 16% Pacific, and 11% Maori [ and Peter Robinson of 3.125% maori is classified as Maori for these stats ].
        So, why is Auckland stated to be an ‘indigenous city of the Pacific’, rather than a westernised Asia-Pacific city? The NZ Herald didn’t print my complaint letter, perhaps because I had enough grasp of reality to also note that Sydney Australia has about 15 % of its inhabitants with Irish ancestry, so this means Sydney is then an Irish city of the Pacific.

            1. I wrote another two paragraphs in my original post that doesn’t seem to have reached the blog, so here they are :
              Some time ago I emailed Dr Lipson of the NZ Initiative about these observations of educational decline. She didn’t reply to my query, which basically noted that of all ethnic groups, school pupils of East Asian ethnicity, irrespective of the host nation, seem to have achievements in science, maths and reading that are relatively impervious to differences in family income/wealth compared to all other ethnic groups. Therefore, the educational performance of ethnic East Asians in schools can serve as a proxy, across time and different national curricula, for the quality of educational input. The UK has very high quality data about the educational performance of ethnic subgroups since it collects this data eg for ethnic Chinese, South Asians from Pakistan versus India versus Bangladesh versus Sri Lanka etc.

              The PISA data for Singapore is basically 80% the performance of ethnic Chinese, and for Hong Kong it is proxy for 98% ethnic Chinese. On the other hand, NZ is distinct from other comparator Western nations in having a school population that has an exceptionally high % of non-Asian non-Whites. The Pacific Island median age in NZ is 25, and for Maori around 27, while White median age is much higher. This means that the % school population in NZ who are Maori and Pacific Island is much higher than their amounts in the total national population : 2018 census showed of 5 M population, 11% PI and 16% Maori. The % of school age Maori and PI students has burgeoned since 1995, that is, at about the same time the national PISA achievements started to decline against comparator western and East Asian nations. So the critical information is hidden in the NZ PISA etc statistics. If one splits the NZ PISA data across time into three cohorts – Maori students, PI students and Asian students, this will actually get into the heart of the matter.

        1. Excellent post Ramesh, thanks. It was very interesting to read your take on things. I hadn’t heard of the NZI at all, but having looked at their site I’d agree with your take – basically a right wing wank-tank pushing a particular ideological position. Doesn’t mean they’re wrong in this case, but it makes me suspicious.

          Re educational decline in general, it seems as if people have been banging on about declining educational standards for ever. When I went to school in England in the 1960s, “modern maths”, or “new math” for US readers, was ruining maths education. Tom Lehrer wrote an entertaining song “New Math” in 1965. Before that the literary critic FR Leavis used to bemoan the declining standards since the days when he used to read Milton in the trenches during World War 1 and everyone apparently could do Greek composition at the drop of a hat. Then from the late 1960s through the mid 1970s we had the Black Papers on Education, pointing out the inevitable ruin that would be caused by comprehensive schools. A bit later, much to my surprise, the period when I was at school was briefly viewed as a golden age from which there had been a steady decline. No doubt there were people in the late 19th century bemoaning the fact that Euclid’s Elements was no longer taught in English schools.

          None of which means that Briar Lipson is wrong about the last 20 years being uniquely bad, although I rather doubt it. I’d want to see the tests they’re using and see what some one like Andrew Gelman makes of their methods to be sure it’s not just another one of the bogus studies like himmicanes or pizzagate or the sleep guy one that he likes to poke fun at. But it seems a bit of a distraction from the main point, which is that the teaching of mātauranga Māori as science in science classes is definitely a bad thing. I worry that dragging in a wide-ranging critique of educational standards over the last 20 years, especially from a source like the NZI, is only going to make it easier for the usual suspects to ignore or pour scorn on people genuinely worried about the specific issue as just a bunch of right-wing nutjobs, and probably Trump supporters to a man or woman.

          Actually, it wonder if Tom Lehrer could be persuaded to do a song on mātauranga Māori. I believe he’s still with us.

    1. People like John Dewey. It has existed in one form or another for close to 100 years. It is extremely prevalent – especially in North America. (A good – very critical – overview of the history of it can be found in books like E. D. Hirsch’s “How to Educate a Citizen”).

    2. Child-centered learning, if done right, is a very valuable approach. It doesn’t work unless the teaching environment is set up for it, and the teacher(s) are trained.

      I gather from what is presented above, that the (incorrect) definition of child-centered learning is to just let the kids do whatever they want. The real deal involved taking the lead from children’s interests and incorporating that focus into the subject matter. It’s difficult to do when done right because there is a lot of individual attention that has to be paid by the teaching staff, and many schools don’t have a small enough student-teacher ratio to accomplish it.

      As an example, when introducing kids to a study of the natural world, some kids might be interested in insects, some in mammals, and some in microscopic organisms. So, I’ve named three possible starting points, and from there the teaching can radiate out. But if you just start with microscopic organisms, you might see the other kids lose interest, when it doesn’t have to be that way.

      I speak from personal experience, too. My intro to biology in high school was mind-numbing. I hated it. I didn’t care about worms and crayfish; I cared a lot about observing mammals in nature. I was in an environment (a lake community) that contained otters, and I would borrow a neighbor’s canoe, paddle out to where the otters were, and watch them for hours. Imagine if that interest could have been incorporated into my studies. But no, it was one-size-fits-all. It wasn’t until I was an adult that my interest in biology was rekindled, and again, it started with mammals, only this time it was horses, and then goats. I learned about nutrition, about poisonous plants and reptiles, and about reproduction, and have continued to learn for my whole life.

      Please don’t throw the baby out with the bath.


      1. Is it something similar to Montessori school? If so, my understanding of that is it is small scale and not likely for the general public.

        1. Yes.

          Maria Montessori herself said, “Let the child lead.”

          And you’re right, it IS small scale.

          It’s also not for every student. And that’s another problem with public schools. Learning styles can vary widely, but the desks-facing-forward-teacher-in-front-lecturing is, again, one size fits all. And in large classrooms, it can be a necessity.

          Hands-on learners like me get suffocated. And, I’m not alone. I have a couple of friends whom I met in elementary school who felt the same way I did. Couple that with the attitude that, “The smart kids will get what they need”, and the result is that lots of talent gets squashed. I looked for a PhD. program that would accommodate my learning style, and the whole time I was in the program I kept wondering where and who I’d be if I had had that sort of learning environment from the beginning. Also, when I showed my dissertation proposal to several of my friends who were in more traditional programs, almost to an individual I got the same response – “Wow, this is really cool. You’d never get to do this where I am.”


          1. A while ago, I proofread and indexed a book that touched briefly on the need for university educational settings to be flexible enough to cope with different teaching methods, which it referred to (if I recall correctly) as “the sage on the stage” versus “the guide by your side”. (The book was called The Physical University and was made up of separate chapters on different aspects of university campuses including the architectural, environmental/ecological, and philosophical amongst others.)

  2. > while NZ has dropped in reading, science … the percentage of students leaving school with their NCEA certificates at level 2 or better (1) has grown from nearly 60% to 80%

    One way to read this is that 20% of the class who used to drop out, are instead sticking around, and dragging down the average. Is this ruled out somehow, for instance by looking at how the top quarter performs? (I did not read all the links.)

    The “year 5” results (at which age I presume everyone is in school) don’t show so much, maybe noise.

    > 32% real rise in per-pupil spending

    I also wonder if this is just that inflation in service-heavy sectors is higher than the average inflation, which includes material-heavy things.

      1. My point there is that inflation adjustment is a tricky thing. At least in recent decades, the prices of things involving human service have gone up much faster than other things:

        Which I think means we should not necessarily read increasing spending as an attempt to do better, to prioritize education. Trying to buy the same thing as before will cost more, in “real terms”, meaning compared to some averaged basket of goods & services.

  3. The implication is that for some decades, New Zealand education has been very “progressive” in a now old-timey School of Ed sense—“child-centered” theory and all that. Another hint in this direction: teaching NZ children to read has apparently underemphasized phonics for decades (see ). My guess, without other information, is that teaching of arithmetic in NZ has not emphasized concepts like fractions, notorious for not being “child-centered”. It would seem that the new fashion of teaching matauranga Maori as if were Physics is just the latest manifestation of this set of educrat foibles.

    I have friends who independently emigrated from the US to New Zealand: one in the 1980s and one couple in the 1990s. Both initially reported some delight in encountering a culture that seemed to be
    decades behind the US or Britain in certain respects. I wonder if the NZ educational establishment has
    in fact been stalled in a collection of educrat fads that were fashionable decades ago in other parts of the Anglosphere. BTW, the friend who emigrated in the 1980s became, after a decade or so, almost
    manically disillusioned with his particular academic situation there, and fled in desperation to the UK.

    1. One thing to be said about the UK is that there isn’t a single education system for the whole nation. Each of the constituent countries has a system that is independent to a greater or lesser extent. This is why the first graph uses “England” rather than “UK”: it’s the results only for England.

  4. “coequal” by the way is pretentious jargon. Equal suffices. One cannot be unequally equal or equal one to the other or whatever pretentious jagon word the ignoramuses invent.

    Of course I know you are only quoting.

  5. Has anyone actually had an in-depth discussion with any of these woke educators to try to find out what their long term expectations of the results of their policies are?
    Lowering educational standards and expectations seems like it is only going to widen the divide between high and low achievers. The rich are not going to be affected. The less affluent people who really value education are going to find a way to educate their kids no matter what. It is the people who depend on publicly funded schools who are going to suffer the effects of this. Education is the most dependable strategy for getting out of poverty. Every bit of lowered standards results in some number of poor and middle class kids who will not be prepared for higher academic disciplines that they would otherwise succeed in.
    Well, that is my opinion, any way.

  6. On the meaning of NCEA levels, there was a bit of confusion around these over at Richard Dawkins’ site which was nicely cleared up by a commenter:

    “ Just add to some clarification regarding education levels, as people are confusing curriculum levels, NCEA levels, and NZQF levels…

    The curriculum levels are from 1 to 8, with year 11 being level 6. They obtain NCEA level 1 qualifications at this level.

    Year 12 being at curriculum level 7, and NCEA level 2 being awarded.

    Finally year 13 working at curriculum level 8 and NCEA level 3 being awarded.

    Then you have the NZQF levels (which is what I think JoMate is taking about), ranging from 1 to 10, 1 being NCEA L1 and 10 being a doctorate.

    It may sound complicated but the UK has similar with key stage levels, GCSE, A-Levels, and it’s own qualification levels.

    Hope that helps.
    (UK trained Math teacher living in NZ/Aotearoa)”

    Students normally start year 1 at age 5, so according to the above NCEA level 2 would be around age 17, with level 3 involving a further year at age 18, presumably for students intending to go on to university.

Leave a Reply