Te Reo - Wai 262 Analysis of State of Te Reo & Critique of Surveys

Sections on the Measurement of the Health of Te Reo Māori

(Excerpts taken from WAI 262 EARLY REPORT)

5.3.4 Censuses and surveys

(1) Pre-1996 national speaker estimates

We have already noted the findings of Richard Benton’s 1970s survey on the health of te reo, especially the scarcity of fluent speakers among Māori children.

The reo which Benton measured as ‘fluent’ in the 1970s was probably at a higher level than that considered fluent today, given that there were many more older native speakers of te reo alive then. As Māori language academic Ian Christensen has remarked,  ‘A tendency towards a diminished perception of fluency may be a natural characteristic of a language in decline.’133 

In a 1992 report commissioned by the Ministry of Education to engender discussion on a New Zealand languages policy, Dr Jeffrey Waite projected the results of Benton’s survey forward to 1986 with corrections for mortality and other demographic variables. This showed that, at the time of the Tribunal’s report on the te reo Māori claim, only 700 North Island Māori children under the age of 10 were fluent in te reo, as opposed to 19,400 fluent speakers aged 55 and over. This did, however, appear to represent an increase in the number of younger speakers from that estimated by Benton in 1979. Overall, Waite guessed there were 81,000 fluent and marginal speakers of Māori in the North Island in 1986.134

In 1995, Statistics New Zealand conducted a national Māori language survey on behalf of T e Puni Kōkiri and Te Taura Whiri. It confirmed Benton’s conclusion that te reo was in a perilous state, finding that 8.1 per cent of Māori aged over 16 had a high proficiency in spoken Māori, 51.3 per cent had low to medium fluency, and 40.6 per cent had no proficiency. Put another way, it showed there were just over 22,000 highly fluent adult Māori speakers – a significant decline from the 64,000 revealed by the 1975 survey. Nearly three-quarters of those highly fluent were aged 45 and over.135 Obviously, if children learning at kōhanga and kura kaupapa had been included, the figures would have been somewhat different.

(2) Māori-language education demand surveys, 1992, 1995

Two surveys conducted in the first half of the 1990s indicated the then potential market for Māori-language education. The first survey, conducted in 1992 by AGB McNair for the Ministry of Education, canvassed the caregivers of 500 Māori and 500 non-Māori pre-school and primary school children and suggested that supply was a long way off meeting Māori demand for Māori-language education.

According to the survey, some 77 per cent of the caregivers for Māori children wanted their charges to receive at least some primary-school teaching in te reo (over and above learning Māori as a subject), but only 33 per cent of those with school-age children had their children in such schools. And, though a mere 7 per cent of caregivers wanted their children to have little or no Māori language taught, 50 per cent of school-age Māori children were receiving just this kind of education.136

At that time, there were 89,115 regular classroom Māori primary-school students but only 13,671 Māori students in Māori-medium classes at primary school (15.3 per cent). This is a far cry from the more than 68,000 that would have been seen if the preferences of the 77 per cent of caregivers had been met. Seventy-seven per cent of Māori caregivers also preferred that their children receive Māori-medium education at secondary-school level, though because those children had not yet begun secondary school, we have no placement figures to compare with those preferences.

However, at that time, there were 37,061 regular classroom Māori secondary-school students, and if a 77 per cent demand had been met, there would have been 28,537 students in Māori-medium classes. Instead, there were just 2,380 (6.4 per cent).

Non-Māori children tended to be much more likely to attend a type of school that accorded with their caregivers’ preferences. Notably, 7 per cent of the caregivers preferred their children’s primary schooling to be in Māori and English, with 2 per cent preferring their instruction to be mostly in Māori.

While the survey indicated that this ambition was met for most of those who held such preferences and who had children already at school, 137 the percentages are more important than they first seem, because if accurate they would have translated nationally to a relatively significant number of children (that is, 29,546 students out of the 328,286 regular classroom non- Māori primary-school students).

In actuality, the number then in Māori-medium classes was 1,275 (0.4 per cent). Māori-medium education at secondary school level was also preferred by 9 per cent of non-Māori caregivers, which would have translated into 17,177 of the 190,851 regular classroom non-Māori students in Māori-medium classes. The real figure was just 100 (less than one tenth of a per cent).

The second survey was carried out for the Ministry of Education by MRL Research in 1995 and was intended to ascertain the likely demand for Māori and Pacific Island language education to 2020. Accordingly, 650 Māori and 550 Pacific Island caregivers for children aged 10 or under were interviewed in Auckland and Wellington.

The results were similar to those recorded in 1992, in that a 68 per cent demand for Māori-medium primary school education was being met by a 43 per cent supply, while a 14 per cent preference for education weighted most heavily towards English was contradicted by a 39 per cent placement in such schools. The trend continued at secondary-school level, with bilingual learning wanted by 57 per cent of caregivers, instruction mainly in Māori by 5 per cent, and Māori immersion by 4 per cent.138

There were then 97,091 regular classroom Māori primary- school students. Had 68 per cent of them been in some form of Māori-medium education, there would have been 66,022 such students, but the figure was only 19,044 (19.6 per cent). At secondary-school level, there were 38,049 regular classroom Māori students, and had the preferences of the 66 per cent of caregivers been realised, there would have been 25,112 in Māori-medium learning. Instead, there were 2,943 (7.7 per cent).139

 At both levels, the clear gap between supply and demand again seems irrefutable.

These results allow us to comment on the demand for Māori-language instruction in the 1990s and the extent to which that demand was being met. While margins of error exist and there is some evidence of slightly reduced demand and somewhat improved supply in 1995, there is a striking consistency across the two surveys. It is, of course, unknown whether places in such forms of education were full or whether a large number of Māori students had Māori-medium learning options available locally but were not making use of them. In other words, the rate of placement cannot be regarded simply as the rate of supply. 

However, given that there was a shortage of Māori-medium teachers at the time, it is unlikely that the actual level of supply was significantly higher.  Even if we assume that the surveyed level of demand was exaggerated, 140 this would not bridge the clear chasm between supply and demand. For example, if the actual level of demand in 1992 was radically lower – say only 35 per cent instead of 77 per cent – this would still have meant that 17,500 Māori primary school children were not attending their caregivers’ favoured form of Māori-medium education.141

Overall, one can thus see that the supply of Māori medium schooling probably improved between 1992 and 1995 but that Māori demand, while still high, may have fallen slightly. Peak demand (in terms of the proportion of Māori students in Māori-medium learning) came in 1999. In the decade since, demand has clearly declined, irrespective of supply, although of course we must remember that ongoing teacher shortages have shown an incessant supply side problem.

(3) Census results, 1996–2006

The 1996 census was the first to ask respondents which languages they could hold a conversation in about a lot of everyday things. It found that 25 per cent of the Māori ethnic group could hold such a conversation in te reo Māori. Nena and Richard Benton found this an ‘amazing revelation’, having assumed, on the basis of the 1995 national survey, that the result would be far worse. Half the speakers were under 25, whereas the 1995 survey had suggested the median age of speakers aged 16 and over would be closer to 50.142

There have now been three censuses asking a language question, and further significant Te Puni Kōkiri commissioned surveys into the health of the Māori language in 2001 and 2006 (see below).  Setting the results of the 1996 census alongside the 2001 and 2006 results, we can discern medium-term trends in the health of the language (see table 5.9)

• The proportion of those aged zero to nine who can speak the language has declined significantly since 1996. 
• In all the age groups from 10 to 39, the proportion of te reo speakers rose between 1996 and 2001; for the 10- to 24- and 35- to 39-year-olds, this proportion declined again by 2006 (in the case of 10- to 19-year-olds to less than 1996 levels) ; and for the 25- to 34-year-olds it continued to climb, but at a much reduced rate. 
• For the 40- to 64-year-olds, there was an ongoing decline which was dramatic at the older levels (for example, from 47.8 per cent of those aged 55 to 59 in 1996 to 33.2 per cent of those aged 55 to 59 in 2006). 
• Amongst those aged 65 and over, there was a marginal decrease in 2001 and a steep decline in 2006. 
• In 2006, the age groups with the lowest proportions of reo speakers were those spanning the years zero to 14. As these also happen to be the most populous, the more positive responses – such as the nearly 50 per cent of those aged 65 and over who were speakers – represent much smaller numbers of people.
• The key concern about this lower-speaking ability amongst the young is that it was not the case in 1996, when those aged zero to nine had higher proportions of speakers than those aged 20 to 29, and those aged 10 to 14 out-rated those aged 20 to 34. 

While the reasons for these changes are undoubtedly complex, some trends do seem readily explicable. 

• The decline in younger speakers would clearly seem to relate to the drop-off in those attending kōhanga reo and the declining proportion of those attending Māori-medium schooling. 
• Conversely, the rises among some age cohorts will relate to factors such as the increased participation in Māori-medium schooling in the late 1990s or the growth in those in later age brackets taking tertiary courses in te reo (notwithstanding Earle’s comment that such courses would not enable one to converse proficiently in Māori on their own).143 
• An example of the latter may be the 30- to 34-year-olds in 1996, who as 35- to 39-year-olds in 2001 and 40- to 44-year-olds in 2006 increased their proportion of reo speakers. 
• The decline of speaker proportions in the older age groups also clearly relates to the fact that, as many older speakers pass away, they are increasingly replaced by those who have never learnt te reo.

(4) Projecting the census results forward

Looking to the future, we know roughly how the Māori population pyramid will look in 16 years’ time. 

By 2026, according to Statistics New Zealand, the Māori population is likely on mid-range projections to number 811,000 – up from 624,000 in 2006.144 It will be older, but still have a larger-than-average number of younger people.145 

If current trends continue, and the proportion of children aged zero to four able to speak Māori continues to decline across censuses, we estimate that (and using an approximate analysis based on the ageing of current age cohorts): 

• 16 per cent of the 258,000 Māori in the zero to 14 age range will be te reo speakers in 2026 (unadjusted for those too young to speak) will be te reo speakers in 2026.  
• 20 per cent of the 303,000 aged 15 to 39 will be te reo speakers in 2026, 
• 24 per cent of the 181,000 aged 40 to 64 will be te reo speakers in 2026, and 
• 26 per cent of the 69,000 aged 65 and over will be speakers. 

In other words, it is unlikely that the official tally of Māori speakers of te reo Māori in 2026 will be more than 150,000.146  That is a rise of 14 per cent during a period in which the Māori ethnic group population is projected to rise by 30 per cent (on medium projections). 

The estimated number of speakers represents a likely 20 per cent proportion of the official 2026 census-night tally for the Māori ethnic group, compared with 23.7 per cent in 2006. It is also likely that, by 2026, there will be very few older native speakers of te reo left. 

Today, those with higher degrees of language proficiency are found in the older age brackets. It is unlikely that the overall proficiency of those 150,000 speakers in 2026 will be any better, if better at all, than the 131,610 Māori speakers of te reo today. Current trends, therefore, suggest that the ongoing gains being made with te reo are not offsetting the ongoing losses occurring as older speakers pass away. 

Moreover, the theoretically ongoing gains are in fact beginning to turn into losses amongst the crucial younger age groups, who represent the future health of te reo. 

In its report on The Health of the Māori Language in 2001, Te Puni Kōkiri stated, with respect to the census results, that :

The predominant feature between 1996 and 2001 is the stability of numbers of Māori speakers at all levels; there is even some moderate growth in some areas. This suggests that the long-term decline in the number of Māori speakers that occurred over a number of decades may have been arrested.147

When the 2006 census results were released, officials suggested that the small increase in the number of Māori speaking Māori represented a stabilisation of te reo after a long period of decline, with a likely rise in the number of younger speakers. 

In fact, however, the age group recording the biggest growth in te reo speakers between 1996 and 2006 (in absolute numbers) was those aged 60 and over, as the population aged. Speakers in this age group increased from 13,647 in 1996 to 16,095 in 2006.148 By contrast, the numbers of speakers aged zero to 14 declined from 38,595 in 1996 to 35,151 in 2006. 

The 2006 result does not appear to be evidence of a further stabilisation at all.

 (5) Te Puni Kōkiri’s 2006 survey

In 2006, Te Puni Kōkiri conducted a survey on the health of the Māori language that seemed to contradict the census result. Announcing the results in July 2007, the Minister of Māori Affairs said that they showed ‘significant progress towards the achievement of the goals of the Māori Language Strategy’. 148 He said highlights included:

• a 9 percentage point increase since Te Puni Kōkiri’s 2001 survey in the number of Māori who could speak more than a few words and phrases (that is, from 42 per cent in 2001 to 51 per cent in 2006) ;

• a 7 percentage point increase in those who could speak te reo very well, well, or fairly well (that is, from 20 per cent in 2001 to 27 per cent in 2006) ;

• the numbers who could understand (by listening), read, and write more than a few words and phrases increasing by 8, 10, and 11 percentage points respectively ;

• the number of 15- to 24-years-olds who could speak te reo increasing by 13 percentage points and those 25 to 44 by 16 percentage points ; and

• an increase in adults speaking te reo to their pre-schoolers at home by 17 percentage points, to primary school children by 14 percentage points and to secondary school children by 20 percentage points.149

As noted, the Ministry of Education also hailed the survey results, arguing that increased Māori proficiency in te reo since 2001 had been helped by the substantial growth in enrolments for tertiary te reo Māori courses during that period.150

(6) Discrepancies between the 2006 census and survey

The 2006 census and the 2006 survey are thus at odds with each other. While T e Puni Kōkiri found major improvements in speaking proficiency amongst those aged 15 to 44 since its previous survey, the census showed declining proficiency among those aged 15 to 24, a very marginal improvement for those aged 25 to 34, and a decline for those aged 35 to 44.

The very small improvement in speaking proficiency for those aged 55 and over in the survey contrasts with a major decline amongst those in this age group in the census. Te Puni Kōkiri has publicly stated its view that:

The Māori Language Survey is a better measure of the Māorilanguage [than the census] as it is a face-to-face interviewand has a variety of questions that investigate languageacquisition, skill and use. It asks a number of questions, eachtargeted at an aspect of language revitalisation that we needto know about.

This survey provides a more robust way to look at thehealth of the Māori language than a single question whichrequires a large degree of interpretation.151

Despite this, the Ministry of Social Development’s influential Social Report for 2007 was equivocal about whether progress was being made or not. It noted that the survey and census data were ‘not directly comparable’ and concluded that:

The 2006 Census shows a slight decrease in the proportion of Māori who speak Māori since 2001, while the 2006 Surveyon the Health of the Māori Language shows an increase overthe same period. It is not clear whether the proportion whospeak Māori has declined slightly or increased.152

Dr Peter Keegan from the School of Māori Education at the University of Auckland has also commented on the 2006 census and survey results, saying that the question of whether te reo Māori ‘is gaining or losing ground today’ was ‘difficult ’to answer.153

Linguist Dr Winifred Bauer of Victoria University has conducted a comprehensive comparison of the census and survey results for 2001 and 2006, and is less than impressed with the reliability of the 2006 survey.

She argues, first, that changed sampling methods and reportage of data between the 2001 and 2006 surveys make ‘serious survey comparison impossible’. She then points to the large margins of error in both the surveys (particularly when focusing on small subgroups within the overall survey sample), which were even bigger in the 2006 survey. She also notes the added potential for unreliability in the 2006 survey introduced by the particular sampling method.154

Most importantly, Dr Bauer says that the 2006 survey is simply not credible because it is so at odds with the census results in respect to general speaking proficiency, the gap between men and women’s proficiency, and the use of te reo by children. Many of the gains claimed by Te Puni Kōkiri relate to very small numbers of survey respondents, are well within the margin of error, and are achieved by combining those stating they can speak ‘very well’ and ‘well’ (since the former group is too small on its own for any credible analysis). By contrast, the census has asked the same question of the entire population, so there are no sampling errors and the results are directly comparable.155 

Overall, Dr Bauer concludes that :

The surveys simply do not tell us what lies behind the key trends discernible from the census, and in fact ‘have failed to provide a better picture than the censuses in crucial areas’. Consequently, it is arguable whether these five-yearly national surveys ‘have any value’.156

The survey results contradict reality: that the health of the language continues to decline. Certainly, there was no improvement in the language proficiency of the critical parenting generation cohorts, who are vital to intergenerational transmission, between 2001 and 2006.157 There is real danger in casting the 2006 survey results in such a positive light. Doing so will encourage complacency about the health of the language at a time when a sense of urgency is still needed.158

5.3.5 Conclusions : how healthy is te reo in 2010 ?

There was a true revival of te reo in the 1980s and early to-mid-1990s. It was spurred on by the realisation of how few speakers were left, and by the relative abundance of older fluent speakers in both urban neighbourhoods and rural communities. The revival was a Māori movement, it was achieved through education, and it was incredibly successful at a grass-roots level. The movement was perhaps at its most powerful during its earliest surge, as  demonstrated by Māori born from 1977 to 1981 being more likely to speak te reo than those born either from 1967 to 1976 or from 1982 onward (see table 5.10).

From around 1994 to 1999, te reo has been in renewed decline. The problem is not just one of declining numbers of Māori speakers but also, strikingly, declining proportions, for it has also coincided with a significant rise in the number of younger Māori. Critically, the decline is now occurring at both the young and old ends of the spectrum. The figures clearly contradict the perception that, among Māori under 40, it is younger people who are more likely to speak Māori. The figures also show that the most populous Māori age groups are also the least likely to be Māori-speaking (see table 5.9).

All this means that, if trends continue, over the next 15 to 20 years the te reo speaking proportion of the Māori population will decline further, even as the absolute number of speakers continues to slowly climb. And despite the higher numbers of te reo speakers likely to be found in, say, 2026, they are likely to be less fluent than speakers now, given the relatively few older native speakers who will still be alive.

The 2006 Te Puni Kōkiri-commissioned Māori language survey showed much more positive results than the 2006 census, but it has been strongly criticised by a leading scholar for its lack of reliability. 

The survey certainly does have large margins of error. Moreover, its inconsistency with key trends apparent in the census and backed up by other data sources suggest it is unwise to proclaim, as did the Minister of Māori Affairs, that the results showed ‘significant progress’ towards achieving the Māori language strategy goals. Needless to say, the decline in te reo overall – and in particular the loss of older native speakers – must be having a major impact on the health of tribal dialects. 

By definition, older native speakers are speakers of dialect. This by no means holds true for children today whose first language is Māori. Something of the fate of tribal dialect is indicated by the fact that there were 20,190 Māori te reo speakers born before 1942 in the 1996 census, but only 11,031 speakers of the same cohort in 2006. 

By 2026, there will probably be not many more than a couple of thousand. In certain areas of the country, of course, the loss of older native speakers is more pronounced than elsewhere, as shown by Te Puni Kōkiri’s regional profiles of the health of the Māori language. In any language with faltering health – or, in this case, a faltering revival – its own variations must be its most vulnerable elements. This is the inevitable state of tribal dialects today, with some elements already all but gone and others clearly in peril. Unless dialects begin to be spoken more by younger Māori, their prospects beyond the next 20 years are obviously bleak. 

The current decline in te reo Māori seems to have several underlying causes. They include:

• the ongoing loss of older native speakers who have spearheaded the revival movement ;
• complacency brought about by the very existence of the institutions which drove the revival;
• concerns about quality, with the supply of good teachers never matching demand (even while that demand has been shrinking);
• excessive regulation and centralised control, which has alienated some of those involved in the movement; and,
• an ongoing lack of educational resources needed to teach the full curriculum in te reo Māori.

The issue of teacher supply strikes us as crucial – the 1992 and 1995 surveys showed the potential market for Māori language education, but the amount of Māori medium education available has clearly never come remotely close to those levels. We are unaware of any attempt to follow up on these these demand surveys, which is of itself a concern. We suspect that demand would be less today, highlighting the failure to capitalise on past momentum.

Successes in Māori language education are today confined to pockets. Undoubtedly, excellent speakers are coming through kura kaupapa and wharekura, but this does not offset the overall decline in Māori participation in Māori-medium education. The Ministry of Education wishes to increase Māori participation rates in early childhood education, but would appear content for this increase to be in centres that are typically English medium.

At tertiary level, more students are studying te reo than in the 1990s, and this may be contributing to language revival at some levels. But it will not help produce the teachers so sorely needed while so many te reo.