Ancestry, Birth and Culture

In the week that I have been reflecting on the knotty questions of who we think we are and why, the Australian Bureau of Statistics has released the preliminary findings of its 2011 National Census.

Last year, I wrote this piece about my difficulties with Question 18.

The report on its analysis entitled Cultural Diversity in Australia does little to reduce the concerns I expressed then.

The report reiterates the official view that Ancestry is not necessarily related to a person's place of birth but is an indication of the cultural group that they most closely identify with. It gives insight into the cultural background of both the Australian-born and overseas-born populations when ancestry differs from country of birth.

So how has this been interpreted by 20+ million Australians. The ABS states Just under a third (32%) of people who responded to the ancestry question reported two ancestries. Or to put it another way, more than 68% of people selected one option only. I was one of almost 4.7 million people who ticked "Australian" and moved on.

English was selected by almost 140,000 more people than Australian, but more than half of them made a second choice. Only 38.5% of those nominating Australian ancestry selected a second option. Of the top ten responses, the only groups with a lower multiple selection rate were Indian, Chinese and Greek.

What interpretations can be drawn from these data?

Garbage in - garbage out remains as true as ever it was in the world of data processing.

Unless you make some brave assumptions about how each and every respondent interpreted the instructions, you should take the conclusions with a hefty pinch of salt.

Context is everything.

In Cultural Diversity in Australia, the number of persons nominating Australian ancestry (7098.5 thousand) is said to represent 35.4% of total population (presumably the population of all responses, not people responding). In 2011 Census QuickStats, we read that the second most popular of Ancestry, top responses was [Australian 7,098,486 25.4%]. But, 25.4% of what? Never mind, the explanation under the table is that The most common ancestries in Australia were English 25.9%, Australian 25.4%, Irish 7.5%, Scottish 6.4% and Italian 3.3%. The fact that the percentage of Australian-ness has apparently declined from 29% in 2006 attracts no comment. Pass the salt please.

Pretty graphics can have very little underlying support.

The following screenshot from The Courier Mail of 23 June bears no relation to any number in the published ABS data.

But the fact that our Australian-ness on this measure has increased 6.7% since 2006 must be a good thing.

Perceptions of ancestry are highly individual. To some (including 2xgreat grandmothers, Jane and Agnes) it was a constant in times of great change. To others (like this writer) it is a subject for analysis against specific contextual variables.

Does it have any meaning when aggregated across a diverse population? Can it make a contribution to public policy development? I think that 4.7 million Australians expressed a view on that last August.