The Australian Bureau of Statistics has long been an adornment to policy making and research more generally in Australia: rigorous, independent, trusted, user-friendly. You don’t have to spend too long using the websites of overseas stats agencies to realise we’ve got an enormously valuable and helpful agency in charge of cataloguing our national data. Yes, government budget cuts have created problems in areas like the workforce data, but those of us who prefer our policy evidence-based owe a huge debt of gratitude to the good folk at Belconnen.
But all public service bodies, especially those based in Canberra, are prone to losing touch with the public they notionally serve. Not in some “lazy pencil pusher” stereotyped way, but because it’s almost codified in the way you talk about your work — because it’s a world where “stakeholders” must be “engaged” (there’s even a “Consultation”: line on ministerial briefs), “industry liaison” undertaken, issues identified and “addressed” (especially “threshold issues”), “ownership” taken or imparted, KPIs met, all while doing what you were going to do anyway. Because bureaucratese is designed to create distance, to facilitate non-commitment, to generate the appearance but not the reality of meaningful interaction. Because bureaucrats really are immensely process-obsessed, for reasons that seem eminently sensible when you’re in the thick of it but which seem peculiar when you sit back and reflect on what actual outcomes are being sought. And because of Canberra — a town full of smart, highly educated people that in so many ways is atypical of the rest of Australia, for both better and worse. I know, because I’ve been there, done that, and got the Public Service award, for 15 long years.
That insularity and distance surely played a role in the long-standing ambition of ABS bureaucrats to transform Australia’s five-yearly census from a periodic snapshot of Australia into what is, quite literally, permanent surveillance, in which every Australian will have a file full of their most personal information provided by each census, supplemented by other personal information held by the government such as medical and pharmaceutical records. Our names won’t be on the file — there’ll be an encrypted identifier for it — but as Crikey has repeatedly explained, linking it to an individual will be trivially easy.
Arriving at the belief that this is a good idea is a classic example of insular bureaucratic thinking and the way processes are shaped to accommodate that thinking. In 2005, some luckless ABS official thought it would be smart to commission an independent report on the idea of keeping names to establish unique identifiers for census information. Privacy expert Nigel Waters was duly commissioned and duly panned the idea. This briefly deterred the Bureau, but in 2011 it decided to start the process of establishing ongoing tracking of citizens via a 5% sample from the census that year — without the permission or even knowledge of the “participants”, or any independent assessment of the selection process. It then decided to extend that to the entire population — but this time the ABS made sure there was no risk of it being derailed. Rather than conduct another independent assessment, it conducted an internal review that, quelle surprise, determined the idea was a good one. On the basis of its own assessment, it waited until a week before Christmas last year to sneak out an announcement that it would be retaining names and addresses.
And that announcement failed to actually explain the purposes of keeping names and addresses, beyond the anodyne assertion that it was “to provide a richer and dynamic statistical picture of Australia through the combination of Census data with other survey and administrative data.” There was no discussion of how this the names and addresses — which everyone is hung up on –are just a tool to enable the establishment of a permanent file on each Australian, composed of our most personal details.
This pro forma approach to process, though, has ended up undermining the census itself: in refusing to seriously engage with critics of its plans for lifelong surveillance, the ABS has reinforced “stakeholder” concerns about privacy, and its dismissive attitude and resort to threats of fines for non-compliance — coupled with a seemingly manic Twitter account that began sending out spam last night — illustrate how out of touch ABS bureaucrats are. There are some interesting parallels, too, with the data retention debate: critics of the census are now being accused of putting children’s health or indigenous programs in jeopardy, just as data retention critics were accused of undermining the fight against terrorism and restricting the ability of security agencies to fight crime. It’s policy-by-guilt-trip.
In illustrating the distance between bureaucrats and the community, the ABS confirms exactly the fears about privacy and security that it could have assuaged if it had genuinely engaged with the community on this major change. The result has been to drive many who otherwise readily support the census to refusing to provide their names, or boycotting altogether. ABS staff have gone too far toward seeing Australians simply as a vast data source to be manipulated — by legal compulsion, if necessary — for the purposes of bureaucrats, not as legitimate partners and customers with their own views about the line to be drawn between claims of the public good and personal rights.
I first urged a boycott of the census back in March. Since then, as many more informed and more expert people than myself have emerged to say they, too, are deeply worried about what the ABS is doing, my view has only hardened. The former head of the ABS, Bill McLennan; former NSW deputy privacy commissioner Anna Johnston; former Privacy Commissioner Malcolm Crompton, respected researchers like Leslie Cannold, politicians with a track record of standing up for privacy and against bureaucratic overreach, like Nick Xenophon and Scott Ludlam, public health researchers, epidemiologists, academics who rely on the census but who are mortified at what’s being done. And everyday Australians who don’t understand why the ABS feels the need to place them under permanent surveillance. So I won’t be party to the destruction of my own privacy. I’ll do nothing instead.
Will anti-terrorist agencies be allowed to access the data soon?
What about the RSPCA?
“to provide a richer and dynamic statistical picture of Australia”
That is not only syrupy non-speak, but nonsense. Individual names, like any individual data, do not provide a statistical picture. Let alone a dynamic one, whatever that means (the stats change with time??).
Demographic data, fine. But personal identifiers are not needed for that, as it is the bulk of the data that matter.
There is future danger, as and when the data are linked to security services. We will leave the sneaky, snobby and incompetent Brandis in charge? There is a clear creeping surveillance state under him now.
What price the Liberal Party selling the ABS too, or if not, then pushing it to “monetise” its data? Personal identifiers will be rather juicy at that time…
Well, I assume they want to construct a panel data set – a cross-section that is tracked through time – in which case they do need to be able to link the data from person 862 in 2016 to the data from that same person 862 in 2021. So in that sense they could provide a “richer and dynamical statistical picture”.
Of course, they don’t need actual names for that, but using any other form of labelling really does little to preserve person 862’s privacy in that, as BK has explained, linking to a name would be trivially easy. (Caveat: I am NOT defending the ABS’ practice, just suggesting that there *is* a statistical convenience to what they propose.)
Bravo, Bernard.
I support the census – although only in anonymous form. Figures don’t require names in order to be collated & analysed.
Some of the reasons which, in desperation, have been trotted out in recent days to justify the current census have been risible. Apparently, the figures (and names, don’t forget the names) are vital to indicate to our government where to build a new highway or where to locate an extra school. In reality if a state or federal MP receives feedback from hundreds/thousands of constituents that a road in that electorate from A to B is inadequate for the amount of traffic they do not mainly rely on data from the ABS. In the real world a highway usually materialises when an MP’s seat (from the governing party) is looking marginal. Behold, the promise of a road upgrade or new stretch of highway prior to the next election.
Re schools: there is an inordinate number of medium/highrise residential buildings being constructed in Brisbane’s CBD & around the fringes. This gives a strong visual hint that the current schools will not be adequate to cope with many thousands of new residents. Or do we need a census to alert the government in another 5 years when the residents are already in situ and short of educational services?
We are being told that allocation of government spending is dependent on everybody responding to the census. From my observations spending is apportioned wherever a Treasurer is or isn’t willing to cut from their budget – it’s apportioned with the electoral consequences for their party in mind. Or perhaps I’ve got that wrong….
The census has value but, practically, it’s not the ‘be all’ & ‘end all’ for social planning.
Zut, exaclty as per a comment I made here earlier. If political decisions about hospitals, schools, roads etc was always purely based on need, then yes, that would provide impetus to their desires, but data in this sense is a 5th or 10th order issue, and politics is always the number one factor in decision making.
I too object to the bureaucratic over-reach, the internal review that just happened to provide the answer they wanted but didn’t get the first time, and the fact that they snuck this out in the pre-xmas period shows that they knew what they were doing and that it was never going to be without controversy. They have been sneaky and insular, and it does have real world problems, and surveillance is a huge issue even if we haven’t worked it out yet.
As for giving over tax file records, well there is a social reason for that, which is that we will happily give over that information if it means that everyone else does and it means that everyone has to pay their tax. It’s different, people should be able to recognise that.
Remember the current census had a near death experience in 2014 when ABS said it didnt have the funding and suggested 10 year census period. The 2015 budget gave the ABS a one time $250 million to modernise its data systems and a suggestion it should become self funding. The current proposals are about making ABS data much more valuable to commercial entities to complement their own consumer datasets and increase the amount of funding ABS can get from data sales – charged as cost recovery or not.
I thought it has something to do with this funding issue.
Privatisation sort of?
People are happy to have their name, address, date of birth and Tax File number recorded when lodging a tax return, or obtaining a medicare card, or filing documents with Births, Deaths and Marriages etc. What’s the difference with that and this Census? The Government already has our name.
This year when filing my tax return with the ATO there was no requirement to name who had been overnighting at my residence – nor to clarify our relationship. Haven’t looked at this year’s census form yet but in previous years that personal inquiry/invasion has been required information.
What part of the word Dossier don’t you understand ???
Up to this point we have provided information to each govt department specific to our interactions with them and they kept that to themselves unless you were suspected of being a terrorist.
Now these ABS a’holes are going to link all that information and gather it together into a DOSSIER on each and every one of us. That’s the beginnings of a f…en Police State.
I think you will find that a great many agencies share data and not just about suspected terrorists
Departments have been sharing data at least since the early 1990s, partly to catch the social welfare cheats.