You are here

Roll change and the removal of zoning, 1991-1998

Richard Harker

Zoning was abolished in 1991 with the stated aim of increasing choice for parents. Ministry of Education data show a considerable shift of student populations between schools, leading towards a dual system of higher and middle decile schools for the (mostly Pakeha) middle classes, and lower decile schools for the (mostly Maori and Pacific) poorer classes. To change this, we need a system that provides real choices for families, without the damaging social and educational consequences of the present system.

Journal issue: 

Roll change
and the removal of zoning, 1991-1998

Richard Harker Image Massey University College of Education

The abolition of zoning in 1991 was heralded as a victory for common sense and the opening up of a “market” in education, which would provide choice for parents, thus offering greater equality of opportunity for students. There has been a considerable amount of research since then showing that things are not quite that simple.1

At the time, arguments were made against the removal of zoning, on the grounds that such a move assumed that families were equally resourced to take advantage of any choices that might be available. Well-resourced families could afford to pay for their child’s transport to more distant schools, or even find the fees for private schools. Other families, however, were in no position to afford extra costs; and in thinly populated areas, logistics precluded any real alternative to the local school.

The theory that good schools would grow and poor schools would decline in response to market forces assumed that schools are rather like shops – some will decline and end up like run-down corner dairies, with few goods on offer; others will flourish and grow into supermarkets where everyone can pick up their chosen range of goodies. Unfortunately, quality education cannot be purchased off the shelf or by the metre. It depends as much on what the “shopper” brings with them into the shop as on what the shop has on its shelves. A quality teacher is one who can put knowledge and the learner together, and this requires unique solutions to individual situations – there are no fixed recipes.

In looking at the effects of the removal of zoning in 1991, both Gordon (1994) and Lauder (1994) identified what they called the “spiral of decline”, in which schools at the bottom of the “market” lose students to other schools.


These “lost” students tend to be higher up the SES scale and from families rather more educationally motivated than those who remain. As a result, social problems are concentrated and negative public perceptions of the school are exacerbated. The school begins to lose resources as the roll declines. However, the case studies of schools used to document this “spiral of decline” were carried out in restricted areas and with small samples, hence the need to look elsewhere for indications of the “spiral” at the national level.

Every year the Ministry of Education compiles a mass of data on our schools. These data are made available for use by schools and others for their own purposes. The graphs in this article have been generated from the Ministry data. When looking at the data shown here, it should be remembered that there is no sampling error involved – the data is on all schools. Hence statistical significance is not an issue.

Primary schools

Figure 1 shows the median roll change for our primary schools by SES decile. During a period of demographic expansion in this age group of just over 10%, schools in all decile groups have grown.2 There is a small trend indicating that higher decile schools have grown a little more, on average, than lower decile schools. There is some evidence here for a small degree of SES-related flight, but not a great deal.

However, Figure 2 shows that the large majority of Maori students are in the lowest few decile groups of schools. There are few schools above decile 5 with more than 20% Maori pupils on the roll.

Figures 3 and 4 re-analyse the roll change data for primary schools. Figure 3 covers schools with fewer than 20% Maori pupils, and Figure 4 covers schools with more than 20% Maori pupils.


An interesting difference appears. As Figure 3 shows, between 1994 and 1998, the median roll at high decile schools with fewer than 20% Maori pupils grew substantially. By contrast, as Figure 4 shows, the median roll at high decile schools with more than 20% Maori pupils declined. Overall, there are not many of these schools: 36 in decile 7; 23 in decile 8; 11 in decile 9; and 3 in decile 10.

A reasonable conclusion would be that Pakeha students have moved from high decile schools with more Maori pupils to high decile schools with fewer Maori pupils (unless one wants to argue that Pakeha students have moved to lower decile schools – an unlikely scenario).

Secondary schools

With secondary schools, the roll change situation is dramatically different, as Figure 5 shows. The median roll for decile 10 schools has grown by over 15%. This means that half of the 27 decile 10 secondary schools have grown by more than this. Conversely, the median roll for decile 1 schools has shrunk by over 15%. Parents with resources have moved their offspring out of low decile schools into higher decile schools, with an uncannily symmetric aggregated outcome.



As in the primary sector, analysing the data on the distribution of Maori and Pacific students (Figure 6) shows that they are overwhelmingly concentrated in the low decile schools. These are the schools whose rolls are declining, thus attracting fewer and fewer resources.

If it is Maori and Pacific parents who are moving their children, they are certainly not going to the high decile schools – there are hardly any such students in schools above decile 6. If these parents are moving their children, it seems that they are going to other low (or middle) decile schools that may have a better reputation. Figure 6 indicates that the growth in the high decile schools is largely due to an influx of non-Polynesian students.




These data show that there has been a considerable shift of student populations between schools, and that this shift has a strong socio-economic underpinning, particularly at the secondary level. These data also show that there is a strong ethnic dimension to these changes, at both the primary and the secondary levels. Our national education system, despite much rhetoric about equity and fairness, shows all the symptoms of the “spiral of decline” that Gordon and Lauder documented for Christchurch. Our system now seems to be well down the track toward a dual system – one set of schools for the (mostly Pakeha) middle classes, and another set for the (mostly Maori and Pacific) poorer classes.

What is to be done?

The OECD study on school choice (1994) pointed out some of the counterproductive consequences of parental choice policies, although it also stated that some form of choice is necessary. Fiske and Ladd (2000) provide a comprehensive account of the consequences of policy decisions made during the 1990s in New Zealand, based around market models and parental choice. Their book is subtitled A Cautionary Tale, and is specifically designed for administrators in other countries. It should be widely read in New Zealand by all educators.

From Fiske and Ladd and the data shown here, it seems reasonable to conclude that:

1.&&Decile 1 and 2 schools have become increasingly Polynesian (Maori + Pacific) in their student makeup.

2.&&Low decile school rolls have declined, whereas high decile schools have increased their rolls.

3.&&Students moving into higher decile schools from lower ones are likely to be those with the higher academic aspirations.

4.&&These mobile students are mainly (but not exclusively) Pakeha or Asian, leaving lower decile schools with an increased proportion of Maori and Pacific students and a residual of poorly motivated students.

Taken together, these conclusions point to a growing split among taxpayer funded schools in New Zealand, to the point of becoming a dual track system – one set of schools for children from well-resourced families, and another set for the rest.

The social and educational consequences of the “sink school” phenomenon have been evident for some time, but have not been adequately addressed by government. Indeed, in the early 1990s the government saw this as the logical and desirable outcome of the market. Such schools, the argument went, should close. No one seems to have wondered what would happen to the students or the communities left without a neighbourhood school. These schools can certainly do with more funding, but that is only to treat the symptoms. The underlying causes lie in demoralised communities, with inadequate housing, few prospects for paid employment, and little evidence before them that working hard at school is going to be worth anything in the future.

What is needed is a system that provides real choices for families, without the damaging social and educational consequences that the present system entails. A number of options have been suggested. One is the reintroduction of a form of zoning, with an emphasis on providing choices for disadvantaged groups, which involves decisions on resourcing for transport, supplementary teaching and the provision of a full range of subject options in all schools (or available to all students). This is called horizontal choice, as opposed to vertical choice (our present system), which advantages those who can afford to live near “good” schools or pay for transport to distant schools (or fund their offspring into the private sector).

Two main principles would seem to be necessary for a fair and equitable choice policy in schools funded from the public purse:

•&&Every family should have the right to send their children to their local school confident that they have access to a full range of educational provision;

•&&Over-subscribed schools must use a balloting system to keep their numbers down to an agreed level. Any other approach inevitably leads to social (including ethnic) selectivity.


Fiske, E.B. and Ladd, H.F. (2000). When Schools Compete: A Cautionary Tale. New York: The Brookings Institute.

Gordon, L.. (1994). Is school choice a sustainable policy for New Zealand? In H. Manson (ed), New Zealand Annual Review of Education 4. Wellington: Victoria University of Wellington, pp. 9-24.

Lauder, H. (1994). The creation of market competition for education in New Zealand. Wellington: Ministry of Education.

McCulloch, G. (1990). Secondary school zoning: The case of Auckland. In J. Codd, R.Harker and R.Nash (eds), Political issues in New Zealand education (2nd ed.), Palmerston North: Dunmore Press, pp. 283-302.

Rae, K. (1991). Secondary school enrolment schemes: A case study in policy change and implementation. In L.Gordon and J. Codd (eds), Education policy and the changing role of the state. Delta Studies in Education (1). Palmerston North: Massey University, pp. 107-115.


1&&&There have been some studies of zoning policies by Gary McCulloch (1990) and Ken Rae (1991), as well as research studies that examine the effects of choice on schools subject to this competitive market (Gordon 1994, Lauder 1994).

2&&&In 1994 there were 427,588 five to twelve year olds in school. By 1998 this had grown to 474,046, a difference of 46,458, that is 10.86%.

RICHARD HARKER is Professor of Educational Research and Development, and Director of the Institute for Professional Development and Educational Research, at Massey University College of Education. He has taught and researched for many years in the general areas of Sociology of Education and Research Methods.