Skip to main content


Archived Comments for: The relationship between socially-assigned ethnicity, health and experience of racial discrimination for Māori: analysis of the 2006/07 New Zealand Health Survey

Back to article

  1. Claimed racial discrimination in the New Zealand health system; a rebuttal

    John Robinson, retired

    12 June 2014


    Claimed racial discrimination in the New Zealand health system; a rebuttal

    Dr John Robinson


    This paper ascribes ethnic differences in perceptions of health as due to racial discrimination against Maori.

    “Within New Zealand and internationally, there is recognition of the important role of racism as a basic underlying cause of ethnic inequalities in health”

    “In New Zealand, Māori report experiencing disproportionately higher racial discrimination at an individual level that has been linked to a range of adverse health outcomes, heightened health risk and poorer health care as well as contributing to ethnic health inequalities between Māori and Pākehā (European).” 

    “Results of this study suggest that, in a race conscious society, the way people’s ethnicities are viewed by others appears to have tangible health risk or advantage, and this is consistent with an understanding of racism as a health determinant. Dismantling the structures of racism is complex yet vital in our efforts to achieve a fair society that facilitates equitable outcomes in health and other social indicators and also enables self-determination of priorities and solutions for Māori.” [1]

    It would be highly worrying if this assertion of different treatment based on racial prejudice in the health system were to hold.  It is a serious charge against health professionals.  However, the analysis is faulty and this is not so. 

    The study considered perceptions of treatment and well-being as in answers to questions such as:

     “Have you ever been treated unfairly (for example, kept waiting or treated differently) by a health professional (that is, a doctor, nurse, dentist etc) because of your ethnicity in New Zealand?” ([2] Question 5.10)

    There is considerable uncertainty here of what is measured.  While it is assumed here that the attitudes of health workers determine the replies, it may well be the attitudes of different groups of Maori that are tested, as some will be more ready to perceive poorer treatment based on ethnicity than others.  Given the current debate in New Zealand, with a considerable emphasis on claims of past wrongs and suggestions of a constitution involving separation by race in a partnership model of governance, this is highly likely.

    Such moves towards ethnic separation are evident in this research, which was carried out in the Te Rōpū Rangahau Hauora a Eru Pōmare Maori Health Research Centre.  An assumption of past harm and continuing unequal treatment is clearly expressed in this paper.

    “Internationally, there is substantial evidence of unfair inequalities in health between ethnic groups and, for many countries with histories of colonization, inequalities between indigenous and non-indigenous peoples within the same territory (e.g. New Zealand, Canada, and Australia).  In New Zealand, there are significant and long-standing inequalities in a range of health outcomes, risk factors and healthcare between Māori (indigenous peoples) and Pākehā (European). … Within New Zealand and internationally, there is recognition of the important role of racism as a basic underlying cause of ethnic inequalities in health.” [1]

    The claim of “unfair inequalities in health between ethnic groups”, either now or in the past, with “racism as a basic underlying cause of ethnic inequalities in health” is incorrect.  In fact “more extensive health provision was made for Maori between 1840 and 1940 then has been generally recognised.”  Although much of health funding came from rates, which were not paid by Maori (and this created very real difficulties for many regions), hospitals were open to Maori equally, then as today. [3]

    The two groups whose experiences and views were compared are those who considered themselves to be Maori (“how you classify your own ethnicity”, self-identified ethnicity) and those who thought that they had been recognised as Maori by others (“how other people usually classify your ethnicity in New Zealand”, socially-assigned ethnicity).  Since the socially-assigned ethnicity is based largely on appearance, that group will consist of those with the greater degree of inherited characteristics (most who report themselves as Maori are of mixed ethnicity).  The two groups are similar to the self-reported measures of sole Maori and mixed Maori in the Census, and the reported differences tell a familiar story of a well-established pattern of increased differences with greater Maori identity.  It may be noted that the percentage of those who perceive racism in health treatment (3.4% of those seen as European and 5.6% seen as Maori) are small, with a considerable majority reporting no such perception.  There is no comparison with perceptions of different treatment by other ethnic groups.

    The authors recognise that others have observed differences between differently defined groups of Maori.

    “Other studies have also examined health and social differentials within the Māori population and shown that health and socioeconomic differences exist for different Māori populations based on their self-identified ethnicity.  For example, people who identify solely as Māori have been shown to have more disadvantaged socioeconomic status and worse health than people who identify as Māori and European”

    They do not, however, consider the many possible reasons for differing health outcomes.  Statistics New Zealand, in comments on differentials in life expectancy, points out that the Māori/non-Māori differential partly reflects different rates of diabetes and smoking, as well as socio-economic differences. [4]  Although there were a number of questions concerning smoking in the survey ([2], Questions 3.19-3.27), smoking was not considered as a possible factor here.

    Contributing factors may lie outside the scope of the questionnaire and the full picture is complex.  Most analysts are cautious in reaching a conclusion, as shown by economist Brian Easton in an overview of ethnic differences, where he suggests social contacts (a feature of Maori society, not defined by the actions of others) as a contributing factor. 

    “Econometric work suggests that only one third of the difference between Maori and non-Maori employment participation can be explained by the personal characteristics measured in the population census.”  There may be other personal characteristics not measured, which also have an influence.  “However it seems likely that the most important determinants of the differences are social variables, summarized in the concept of ‘maoriness’. A possible practical example is that it is known that the most important source of job recruitment involves family and friends.  The Maori is handicapped in doing this because of their lower employment rates, but also possibly because the Maori network is not as geared as the non- Maori family to carry out this task.” [5]

    Easton finds that his chosen variables provide a fit for just part of the difference.  He then labels the remainder ‘maoriness’ and seeks a reason for that part of the difference.  The authors here carry out a similar analysis and assume that it is a consequence of racial discrimination.  One problem here is with the understanding of just what is being measured and of the full range of possible contributory factors.  Another is with the validity of the analysis on which they rely.

    The data was analysed to consider whether differences could be due to a limited selection of socioeconomic experiences using “Survey analysis based procedures” and concluded that “In multivariable analysis, Maori who were socially assigned as European-only had a significant health advantage.”

    There are serious limitations to the application of logistic regression.  The choice of independent variables to provide a best fit for a dependent variable may be incomplete, leaving out key causal factors.  The methodology assumes linear relationships across the whole range, which is often far from the case.   And the set of independent variables must not be highly related.  The process can handle a degree of covariance, with some relationships among the independent variables, but when those variables are highly interrelated the process can be unstable and a small change in measures may result in a significant change in the output.  Given the complex and non-linear relationships among socioeconomic experiences (the various social variables such as health, education, income, employment and demographics are clearly interrelated), neither requirement is satisfied.

    Here the whole remainder, left after a questionable analysis for a limited range of socio-economic variables, is labeled racism, ignoring the lack of supporting evidence and the many possible alternatives. 

    The history of New Zealand is of considerable cultural shifts, and many remaining differences are the consequence of past cultural practices.  Maori are descended from a Polynesian people who lost contact with the mass of humanity when they moved away from islands off the coast of Asia to Remote Oceania some 3,200 years ago.  Since the coming of Europeans and others from all parts of the earth they have moved from a Stone Age tribal culture to share the opportunities and life experiences of a modern developed nation. [6]  Those gaps can best be understood by a thorough, robust and comprehensive study of past experiences, social statistics and associated analyses, and not by jumping to a simplistic assumption of racism, in the past and continuing in the twenty-first century.

    This paper makes an unproven claim of racism among health professionals, that “the way people’s ethnicities are viewed by others appears to have tangible health risk or advantage”.  The claim of racism on the part of health professionals is a serious charge that is in no way justified, and not to be countenanced in the absence of definitive proof. 



    1. Harris R, Cormack D, Stanley J. 2013:The relationship between socially-assigned ethnicity, health and experience of racial discrimination for Maori: analysis of the 2006/07 New Zealand Health Survey. BMC Public Health 13:84
    2. Ministry of Health: 2006/07 New Zealand Health Survey Adult Questionnaire. 
    3. Dow D 1999:  Maori health and government policy 1840-1940. Victoria University.
    4. Statistics New Zealand 2009: New Zealand Life Tables: 200507
    5. Easton B: The Maori in The Labour Force.  In, Labour, Employment and Work in New Zealand 1994 (pages 206-213).  Edited by Morrison P S. Victoria University of Wellington.
    6. Robinson J L 2013: A plague of people.  Tross Publishing


    Competing interests

    None declared