Lo and yonder: the 2016 Census is coming in August, but complaints about privacy have begun to hammer through from some more strident libertarian quarters. The criticism stems from the Australian Bureau of Statistics deciding late last year to anonymise and retain this this year’s collection to improve the quality of data available to policymakers. Since then, the Pirate Party have called for a boycott and the IPA’s Chris Berg has concurred, weighing in to warn about privacy breaches and suggest we give the count a skip. However, while these groups drum up concerns about confidentiality, it’s far more likely that any damage related to the census will come from boycotting, rather than the consequences of holding onto the data.
But how did we get here? For the last three censuses, data retention has been an opt-in system, and the public has taken a shine to it, with the 2011 edition seeing 60.6 per cent of people choose to have their data retained. This was up from 56.1 per cent in 2006 and 52.7 per cent in 2001, which indicates that Australians are largely at ease with having their information kept for future generations and for forward planning. The intention is that by keeping data, it’ll be possible for departments to draw on the census record to create better targeted policies for key public welfare departments like health and education. The department’s procedures for keeping the data secure, and their assessments of risk, have been made available and the measures taken are pretty comprehensive: data is separated completely from names, and the collection will not be made accessible until 2115.
In the context of the census as a piece of national infrastructure, trying to raise a boycott doesn’t resemble a reasonable personal protest; it’s more of a distortion, and the ramifications can be severe. After all, to be not counted is to go underrepresented. That might be alright if you’re a North Shore libertarian, but selling a boycott to people who might have needs that require enumeration so services can be better provided tends towards the plain irresponsible. It is no accident that historic undercounting goes hand in hand with historic underfunding; for example, estimates in New York have suggested that undercounting in the 2000 Census was reflected in a loss of funding to vulnerable communities of US$1.2 million for every 100 individuals uncounted in the 2000 census. Getting it right is too important, and being able to draw on accurate, anonymised information across collections provides policymakers with more knowledge to make better informed decisions. Besides, in this age of Big Data, individuals and businesses are bombarded with requests for information (both passively and actively) from governments and other bodies, so the ABS making better use of the information they’ve already received (as opposed to following up a Census with further, targeted surveys) is a bit of a no-brainer, especially during a time of severe departmental budgetary pressure.
While each of us relinquishes a huge amount of data (and our autonomy over it) every time we skip reading the terms and conditions or sign into Facebook, those kinds of data present us as something very specific, and not very helpful to us: a consumer as a subject of knowledge, a form of being known that only benefits us insofar as we enjoy what is marketed to us, and exposes us directly to being manipulated by that information for private ends, for profit. On the contrary, a (politically neutral) census like ours maps the individual and the household as something far different: a citizen with rights, participating in a society that (ought, at least) to make policies to assist them. If there’s data to worry about, it’s the things we’re not aware of that we’re probably best fixed to fret on.
A lot of the fluster comes from the notion of function creep, that even if the ABS and the government were to follow their own rules to the letter, they might later change these rules and exploit the public’s trust. But while the ABS’s data-handling integrity hasn’t changed in over a century, privacy concerns have, and will continue to. In 1905, the main issues were whether people had to disclose their religious beliefs or their income. Besides, much of the information furnished in the census can be accessed on the electoral roll at your local AEC office.
In any case, the pre-2001 time-honoured practice of destroying census documents for security purposes isn’t that time-honoured after all, stretching back to only 1971. It came in pretty haphazard fashion to boot, during a particularly fractious session in the Senate over privacy concerns about a population register card. Then-Treasurer Billy Sweden assured parliament that census documents were destroyed after processing, which turned out to be a bit of a blue – contemporary policy was that the two previous collections were kept on hand for assisting with policy. Panicking, he ordered the destruction of the ’71 census and any other remaining records, which now looks like one of those knee-jerk snap decisions that become part of the furniture once enough time had passed, like driving on the left or laying a table. A tradition which was never tested as part of an electoral platform, it’s been a curious thing to hang on to, especially when Australia is an outlier when it comes to preserving census documents. Canada and New Zealand retain them, records in the USA are available 72 years after the census year, and most European countries hang on to their documents.
Additionally, while we’ve already done well in cutting out a significant archive of information for future generations, we can arrest this. From the vantage point of a hundred years hence, we can learn about the past in forensic and revealing detail, just from examining the particulars of where someone sat on a particular day. Pre-independence Irish census returns are a great example, as the changes in playwright Sean O’Casey’s census documents reveal. Filling out the form on behalf of his widowed mother and brothers in 1901, he reported in English as John O’Casey. By 1911, he was reporting in Gaelic as Seágan Ó Catasaig, and his family had also de-Anglicised. If he’d boycotted the census or the forms had been destroyed, we wouldn’t have this record of how he and hundreds like him reported their resistance to the imperialism of English-language rule. If we don’t hang on to census documents, we can’t even know how people have protested against them.
As this year’s census will permit participants to identify outside of the male/female binary, it may be our most benign census yet, and utterly crucial. It’s a pretty old-fashioned and optimistic thing to believe in, but at a time when our information is harvested at an incredible clip in order to thin our wallets, narrow our attention spans and sculpt our digital environments, providing the government with accurate information to at least make good policy outcomes possible seems like a more reasonable notion than ever.