Editorial: A question to define who you are

What are you? Ask that question of a European New Zealander and he or she will immediately answer, New Zealander or Kiwi. Ask it of a Maori, Chinese or Samoan New Zealander and their instinctive reply might be their ethnicity. When a vigorous campaign urged people to write "New Zealander" in answer to the ethnicity question at the last Census more than 90 per cent of those who did so were people who had ticked the box for "New Zealand European" at the previous Census.

These people were stating their nationality in answer to a question that asked, "What ethnic group do you belong to?". So many answered that way - 400,000, or 11.1 per cent of the population - that Statistics NZ is worried about the integrity of its ethnicity figures and invites comment on the way New Zealanders' sensitivity might be handled at the next count, two years hence.

One obvious idea, add a specific question on nationality to separate it from ethnicity, has not survived a trial run. Given both questions, too many people in a survey still insisted their ethnicity was also "New Zealander". This suggests the write-in campaign was not confusing ethnicity with nationality, as statisticians and academics supposed. The nationalists knew the difference but did not accept it.

They were making a political statement rather than an objective declaration. The way the Census question is phrased, it is hard to blame them. The word "belong" raises emotive challenges that invite a nationalistic response, and "group" is a divisive notion to most people.

Members of the ethnic majority do not "belong to a group". They know the Census question is not about them. It will be used to compare the well-being of minorities with the majority, particularly in health, education and other social services, and it will be a powerful tool in the allocation of public funds.

These purposes are unlikely to endear the ethnicity question to many. Statistics NZ's discussion paper is rather too hopeful that a public information effort might improve compliance with the question in 2011.

Census compilers are reluctant to rephrase questions for fear of making it hard to compare the results with past counts. But the ethnicity question has caused problems for too long. And its wording has been altered over the years in any case, as has the way its results have been classified.

Until 1986 the question inquired about "race" and asked those of mixed race to state their "proportion of blood". That year the concept was changed to one of "cultural affiliation". People were presented with a number of possible identities and asked to, "tick the box or boxes which apply to you".

That seems an eminently better approach than the "ethnic group" terminology adopted later. If the question returns to its earlier format it could offer a box labelled, "New Zealander of European descent", which ardent nationalists might not feel the need to deny.

What is ethnicity anyway? The department defines it as people who share characteristics such as ancestry, religion, customs and language, a community of interests, a common geographic origin. It would be idle to deny that these distinctions often have an impact on the social experience of those concerned. Ethnicity is a valid interest of social study, public policy and providers of goods and services.

But it is equally idle to try to separate it from nationality. Both are about a person's primary identity.

The Census needs to contain a question that invites people to identify themselves by their heritage and culture without threatening those whose nationality comes first.