Axe the 2016 Census? You've got to be kidding!

Canada have done it, with terrible results. The UK have threatened to do it and then gone back. Now it’s Australia’s turn, apparently. The government and ABS is refusing to confirm whether the 2016 Census will go ahead as planned, and is considering options for changing it to a sample survey,  reducing the frequency to every 10 years or eliminating it entirely.

Why would anyone even consider getting rid of the Census?

Cost cutting of course. The Census costs around $400m every 5 years. This is a lot of money. For this we get the most consistent, reliable dataset on the population and housing of Australians, at the finest detailed level, for every suburb, town, locality, local government area in the country. Also every population group by age, cultural group, income level, family type etc.

I don’t know anyone who has NEVER used Census data. It is used at the most fundamental level to redistribute the electorates in this country, a cornerstone of our democracy. It is used at every level of government for planning the essential services needed in our community. Most businesses use it to look at their markets and plan for the future. Academics and universities use it to contribute to debate on all levels of public policies. Students use it for their studies, and individuals use it every day to find out about their own communities.

Not only this, the power of Census is its regularity and consistency over time. Being able to see how we are changing as a nation, state, LGA, town or even a street is incredibly powerful. Who are the emerging groups in our community, are we better or worse off than we were 5, 10, 15, 20 years ago? The Census answers all this.

All this for $400m, or about $20 per person, once every 5 years. In an information economy, this is a fantastic bargain.

Why would anyone cancel it?

$400m is also approximately the cost of the government’s proposed “metadata” program – designed to to allow government agencies to spy on Australians, of no net benefit to anyone. The Census on the other hand is completely confidential, with no personal information kept.

Can the Census be replaced by administrative data and surveys?

The answer is clearly “NO!”

Survey data is all well and good – The ABS runs some excellent surveys providing worthwhile information at the national and state level, but they all suffer from one drawback – the lack of small area data. Survey sizes are generally small, and once disaggregated to LGA level, the sampling error is too high to produce meaninful results.

Administrative byproduct data, such as Centrelink, Medicare data, bank records, utility company databases etc. are often touted as providing an answer to this, and it’s true that you can get some of these at small area level. But they have two major drawbacks:

1. There are enormous privacy issues in releasing this information, which is not collected under the auspices of the Census and Statistics act, and where individual’s confidentiality is not necessarily protected.
2. The datasets all collect small amounts of data, for slightly different populations, and can’t necessarily be directly compared. So the breadth of information and availability of cross-classification for specific communities just isn’t there, and each dataset would have a different degree of reliability.

In short, nothing can replace the Census, and we are fortunate indeed in Australia to have such a fantastic quality Census run every 5 years, to give us a snapshot, not just of the nation but of each place within it. Long may it continue.

To see for yourself the depth of information provided by Census data visit .id’s demographic resource centre.

Read related blogs:

Why we need to save the 2016 Census

Let’s join forces to Save the 2016 Census