Census poses a $38m question

Completed census forms are being collected across the nation. JAN CORBETT and WARREN GAMBLE ask if all this expensive effort is really necessary.

Not many traditions have survived since the birth of Christ as well as the census.

So it seems more than time to ask why we persist with it.

In this electronic information age, when we are watched and counted every time we walk past security cameras, or use our credit cards, loyalty cards or mobile phones, the laborious process of filling out a hand-delivered form every five years seems nothing short of antiquated.

We are already legally required to register births, deaths, marriages and our cars - why can't those statistics be collected from the agencies which already have them?

And even though we are told census results are important for planning health, education and transport systems, for example, why does the supply of those services so seldom match the demand?

A specialist in consumer research, analysis and forecasting, Professor Rod Brodie, of Auckland University's marketing department, says much of what the census asks could be found out from a scientific sample of 1000 people, particularly the "diabolical" question of what energy source you use to heat your home.

Compared with the $38 million it costs to run the census, a sample like that, which businesses often rely on to measure demand, would cost a mere $10,000.

Not that Professor Brodie advocates abolishing the census entirely. Like others in the market research field, he says statisticians do need a quantitative benchmark to measure their sample surveys against. But he says those basic demographic questions would require only one page.

A smaller census form could be scanned more quickly - in an ideal world it could provide a statistical picture of the population next day. And, he says, it would make it easier to take that snapshot more often, which would be more useful.

One of the chief criticisms of the census from the academics who rely on it is the time it takes for meaningful results to be available. Even then, they have a limited shelf-life.

Massey University's Professor Paul Spoonley, who studies migration trends, says cards filled out at airports may tell us who is coming and going and how long they intend to stay or stay away, but only the census will tell us if they fulfilled their intentions.

For instance, he says, in 1996, the date of the last census, the face of and population values in Auckland were radically altered by Asian immigration. Observation tells us many of those immigrants have since left. While he expects to see that confirmed by this year's census, Professor Spoonley admits it is not much use for forward planners to have that population movement confirmed seven years or so after the event.

Even Professor Brodie's colleague Professor Peter Danaher, who says the census results are invaluable for his work with the people meters that measure television ratings, says working with information that is five years old is a problem. It forces him to also rely on the data Statistics NZ collects between censuses.

The Deputy Government Statistician, Dianne Macaskill, says census data are released much more quickly now. In the 1980s, she says, reports were still being produced from the previous census when enumerators were back knocking on doors with the next one five years later.

For the latest census, the national population count will be released in May, and the detailed breakdown reports will begin appearing next February.

The trouble is, says Ward Friesen, a senior lecturer in human geography at Auckland University, there is often insufficient money to pay researchers to further analyse the data, meaning a lot of information that is gathered is never used.

He recalls waiting three years after the last census to get a research grant to use the census results to draw an accurate picture of the ethnic spread across Auckland.

And even if the information is compiled, Dr Friesen says there is often insufficient political will to act on the results.

Remember the population boom hitting Auckland schools in the early to mid-1990s? Dr Friesen, who is involved in education planning, says the population projections had been made. But the Government had abolished school zoning and refused to build new classrooms in Auckland while there were empty ones on the West Coast.

Cheryl Remington, of the Ministry of Education's data management and analysis unit, says census figures are only one of the tools used for planning the supply of classrooms, and that they do gather information from the birth register. But all the academics agree that while the birth register will tell you how many people are born each year, only the census can plot internal migration patterns.

To Ms Remington, one of the best things about the census is that it tells us how well educated our population is, relative to other countries. It tells us, in effect, if the ministry has been doing a good job. And while it is easy to count how many children are in early childhood education, only by comparing those numbers with the census can officials see how many are not.

In the same way that other registers can tell us how many were born, but not where or if they will show up for school, Professor Spoonley says that while there are indeed other agencies which count the number of cars on the road, only the census can tell us how far they travel each day by asking in three separate questions where we live, where we work and how we travel between the two.

Dianne Macaskill, says the census is "absolutely essential" to track population changes and shifts, and enable planning for them.

Census figures are used daily by Government agencies and local bodies as a planning base for hospitals, schools and other services.

In medicine, for example, the census-generated child population figure is needed to work out child cancer rates and uncover trends.

Population figures are also used to revise electoral boundaries.

Locally, community groups also use the information to push local projects. The number of children in a community could be evidence for a local group to lobby the council for a swimming pool.

Dianne Macaskill says sample surveys can accurately project total population, but they cannot detail what is happening in communities.

"If you want to find out whether a new school is needed in Johnsonville and Northcote, you need to ask that community." This, she says, is what the census does.

"Why we have a census is to understand how communities are changing, and there is not an alternative."

A series of sample surveys would be just as expensive and not as detailed, she says.

For example, Statistics New Zealand's own labour force surveys use a 13,000-strong sample to provide regional and national rates of unemployment.

The survey also provides national figures on the age and ethnicity of unemployed, but the sample would be too small to get regional breakdowns of those categories.

Dianne Macaskill says the census has another important function by charting the country's changing face.

Information on ethnic diversity, age and religion helps build up a picture of national identity, who we are now and where we came from.

At $10 a head, she says New Zealand's census is considerably cheaper than many overseas. Countries with considerably larger populations like Britain and the United States run their census only every 10 years.

Other countries use different systems to monitor population changes. In Singapore and Scandinavia, a central government registration system retains personal details and any change of address has to be notified.

But New Zealanders would be unlikely to accept any system where their details could be made available to a number of Government departments. Indeed we have a Privacy Act which prevents that sort of cross-departmental information sharing.

Information from the census here is restricted by law for use only for statistical purposes. Those authorised to see the census forms have to sign a declaration of secrecy.

Rather than abolishing the census, Professor Spoonley argues it should be expanded and improved. Out of concern for accuracy and continuity, he is unhappy about the ethnicity question, even though he was involved in its drafting.

Pakeha was dropped in favour of New Zealand European because some people find it an objectionable term. But equally, a large number of Pakeha would never identify as European.

He says both options should have been provided and other ethnic bands, like Chinese, should be refined to more clearly reflect cultural identity.

And he says we need more questions that show us how the new economy is functioning and how our daily lives are changing.

When New Zealand's first census was held in 1852, it asked how many servants there were in the household, and the number of eggs your hens laid.

Social change eventually rendered those questions irrelevant. But new questions have not kept pace with the rapid social change characteristic of the times we live in.

Why not ask how many of us are hiring outside help to do traditional household chores? How many of us are paying for those services under the table, fuelling the black economy? How many are putting children into childcare? How many of us are regular illicit drug users?

We are assured, after all, that our answers are confidential.

With the expense of running the census, says Auckland University economist Susan St John, after you have read the form it seems like a lost opportunity.

"You'd think it would be possible to put more detail in. We could find out a lot more about what's going on."