Dialogue: A better idea of who we all are

By JOHN ROUGHAN

One of the crucial calculations to come out of any census is the ethnic proportions of the population. Not crucial to you, perhaps, nor to me, but it matters a great deal to those who theorise about society and allocate its benefits.

The whole edifice of social science is built on statistical measures mainly of race, gender and income. Give sociologists some data on illness, unemployment, criminality or whatever and instinctively they will sort it by race, gender and income and compare the percentage in each classification to the percentage of that classification in the population overall.

When they find what they expect, namely that their chosen category of people is disproportionately over-represented in a record of misfortune, they do not quite go so far as to say its distinguishing feature, such as ethnicity, is a cause of the problem. They say it bears a "correlation" with it.

Everyone in the category then becomes a candidate for particular assistance in the form of health programmes, state housing, school equity funding, job training schemes, public sector recruitment quotas and so on.

So it seems important to get the basic population percentages right. The five-yearly Census did not do that when it last counted ethnicity and, although Statistics New Zealand has changed the basis of its published ethnicity figures this time, it still is not giving us a precise picture.

Now, this is not to be one of those diatribes against free ethnic identification. If it is blood you demand, there is a different question on the Census form for you. I am more interested in the ethnicity you feel. Like sexuality, it is so close to the core of who you are that I don't believe you will deceive me.

But I have raised an eyebrow here before about the way Statistics NZ classified people who ticked more than one box in the 1996 Census.

Essentially, the statisticians at that time laid down an order of racial priorities. If you said you were Maori and something else you were counted as Maori. If you said you were a Pacific Island and something else, except Maori, you went into the Pacific Island category.

If your ancestry was Chinese and, say, European, you were counted as Chinese. On it went, down the statisticians' scale of importance, until all that remained were those who identified solely as New Zealand European or Pakeha.

The effect, obviously, was to inflate the Maori percentage of the population and deflate all others, particularly Pakeha. Although Pacific Islanders had second priority, they, too, were heavily discounted. About 25 per cent of Island children born in this country also have Maori parentage.

That count was outrageous. It gave us a picture of New Zealand that effectively excluded the possibility of dual or multiple identity and arbitrarily assigned such people to an ethnic group that might not have been their prior allegiance, if they had one.

The distortion was doubly bad because the ethnic question that year virtually invited people to make a multiple response. The only instruction was to "Tick as many circles as you need to show which ethnic group(s) you belong to."

Nothing on the questionnaire suggested that only one tick would be counted, and it gave no hint that the statisticians had a racial ranking order in mind.

The department did not make its procedure widely known. It was news to me when a demographer, Professor Ian Pool, of Waikato University, criticised the method at a conference two years ago.

Statistics NZ would explain, if asked, that the prioritised count was not its only ethnic data. If you wanted the complete tally of dual and multiple responses it could give it.

But it was the prioritised percentages that added up to 100, and they were the ones of usual reference over the next five years.

Now, thankfully, the 1996 figures are gone, superseded, banished even from historical comparison. Statistics NZ changed both the question and the counting method for last year's Census, and says no valid comparisons can be made with the ethnicity results of 1996.

For the latest Census the department reverted to the question it asked in 1991, simply: "Which ethnic group do you belong to? Mark the box or boxes which apply to you."

It is always a worry to see how much a question can influence the answer. In a report on the Census changes, Kate Long, of the department's social statistics division, notes that relatively few people (42,252) ticked three or more ethnicity boxes last year. In 1996 as many as 131,091 had done so.

But then, even Europeans had been invited to be English, Dutch, Australian, Scottish or Irish that year.

When last year's numbers were released last week, all comparisons given went back to the Census of 1991.

They showed, for example, that New Zealand's Asian population - given relatively low priority in 1996 - has doubled since 1991. They number now slightly more than Pacific Islanders, despite a fairer count of the latter.

In Auckland, says the department's release, one in eight people is Asian and the same proportion are Pacific Islanders.

Maori account for only one in 10. Nationally Maori now make up 14.7 per cent of the population and Europeans have declined from 83 per cent in 1991 to 80 per cent in 2001.

But these figures, too, ought to come with a caution. If all the ethnic percentages are tallied up they would come to more than 100 per cent, because those who gave more than one ethnicity have been double counted or triple counted as the case may be.

It is not clear to me that this is a vast improvement on the fictions of five years before.

If social scientists and public policy makers mean to continue apportioning problems and benefits to population percentages, they need more than ever to remember these are not strictly accurate and the groups are not mutually exclusive.

Each Census, excluding the 1996 rogue, has recorded a decline in the numbers of New Zealanders who identify solely with any ethnic category.

The exception is Asians and the reason is probably that large numbers migrated within the past 10 years.

Kate Long notes: "Recent arrivals in New Zealand are less likely to belong to multiple ethnic groups than people who have been in New Zealand for some time."

Increasingly, our ethnic tapestry has colours bleeding into each other and threads that cross the natural divisions.

It is an encouraging picture really, but you need to look past even the improved statistics.