The first results from the 2001 census were published yesterday. At first blush they confirm some well known trends: incomes are rising; the population is aging; more people are identifying themselves as of Aboriginal descent; Catholicism is the largest religion; more people are living alone; more people are using computers; and Queensland has the highest population growth and Tasmania’s population is declining.
The importance of the census, though, is that it is definitive and that it quantifies the trends. And on that basis, government policy and private-sector decisions can be based. For example, this census included questions on computer usage for the first time. Half of Australian households had a computer and a third of the population used the internet in the week before the census. It suggests that the internet is a fairly good and improving method of information dissemination, but the majority of Australians get their information in other ways. Governments will have to address this information divide. Another example is the 16 per cent increase in the number of people identifying themselves as of Aboriginal or Torres Strait Islander descent since the previous census 5 years ago. Only 12 percentage points of that were natural growth; the rest was change of identity. Governments and indigenous groups will have to address that. Are some misidentifying themselves and therefore taking resources from more deserving or influencing the definition of Aboriginal culture? The census is proof of indigenous disadvantage that is not open to the sort of attacks directed at sampling techniques.
The census provides other important information. Chinese languages have overtaken Italian as the most spoken language at home, indicating Australia’s capacity over generations to absorb new waves of migrants while encouraging them to retain elements of their former culture. The division of population among the states and territories will affect electoral representation and the way federal funds are divided. This census revealed a reluctance by more than 1 million people to reveal their income – presumably at the high end. Combined with fact that Tax Office information gives us only DECLARED income, it might mean the extent of inequality in Australia is disguised.
The Australian Bureau of Statistics faces some difficult choices about how it conducts the census and what it does with the results. The fact this census confirms a lot of well known trends suggests that the questions were not detailed enough. More questions would have elicited hitherto unknown trends and information. However, if people were presented with too many questions, they might have ignored many of them as too intrusive or too onerous.
The question of the fate of the census forms was a difficult one. In an ideal world they would have been kept with full names and addresses for researchers in a century’s time. But if that had been decided, it might have jeopardised the original collection by people fearing (however unjustly) braches of privacy.
And now the information has been collected, the question of its distribution and use is important. The bureau sells a lot of information at commercial rates. There is an argument that the raw datasets at least should be free on the internet to all those Australians who contributed to them. Sure, if bureau staff are asked to do something extra or if printing costs are incurred, the bureau should charge commercial rates.
Importantly, the census is a critical tool for statisticians as a base for later sampling – if the sample’s answers match on census questions match census data, it indicates the sample is sound and the other (non-census) answers are sound.
The census is an important exercise in determining who we are – laying the groundwork for policy-makers to preserve and improve the worthwhile and work on area of poor performance.