Tuesday 9 August 2016

It's 9 August 2016 and at 5:21pm in Australia it will be Where's Wally O'Clock



As public service demographers wait for the personal and collective response to increased sensitive personal data retention to unfold tonight - one last word on #CensusFail from Oecomuse: Musings from home and hearth, posted 7 August 2016: 
There are many valid concerns about the census this year. These include data security and encryption and privacy and data linkages across government agencies and retaining our names and addresses for a longer period (18 months up to four years) than last census, especially when Data Retention legislation has been passed in the meantime.
These issues have been thoroughly covered by far better minds than mine, such as Ross Floate here and Richard Chirgwin here and the indefatigable Asher Wolf and Rosie Williams’ Little Bird Network.
The legal implications of not complying with Census instructions are set out here.
But what is really getting to me are people who say that a robust and accurate Census count will be of great benefit policy-wise, and cite Indigenous health and the homeless population to support their argument.
This is a typical tactic. We saw it when Josh Frydenberg said the superannuation changes would benefit older women. Does anyone really believe that the Liberal Party cares more about older women living in poverty than the pressing need to have just one policy to hang claims about ‘governing for all Australians’ on?
The other typical feature of this argument is that the people making it – demographers, bureaucrats and other academics – are absolutely guaranteed to benefit from data linkages. In contrast, there is no guarantee whatsoever, and not a shred of evidence, that Census data produces benefits for Aboriginal people’s health or for homeless people.
In the ham-fisted way that neoliberal government is done these days, the Census trust problems continue unaddressed. Instead, a glaringly obvious problem is blundered and stumbled through, blasted and blustered at, urgently plastered over, in a make-shift, ad-hoc, amateurish way.
Why? Because the problem is not a significant problem for comfortable middle Australia. It is the tyranny of the majority writ large – a phrase that should be familiar to Liberal Party politicians.
Here are a few of the questions I would put to academic and bureaucratic defenders of the census changes:
Is public schooling a necessity for your children, or a choice? What about hospital cover? Is the GST that a sole parent pays on her child’s school shoes, upfront and at the point of sale, subsidising your investment property? While you judge her? Have you ever been subject to years of Centrelink compliance measures? Have you feared for your life at the hands of your ex/partner?
Do you know anything at all about people in situations you have never encountered, let alone what is best for them?
But science
Why are demographers and other social scientists saying sciency things without any evidence?
Indigenous people on dialysis and with other co-morbidity diagnoses were counted in the last Census. Are they better off? Has anyone asked them? What is the link between data matching and their well-being, whether as individuals or a population?
How is we will know more about your death rates a convincing argument for people who are under more surveillance than any other peoples on earth?
That is not science. It is disgusting and exploitative.
Homeless people were counted last Census too, or at least there was a genuine attempt to reach rough sleepers. But housing is a state responsibility. Has anyone seen a propensity of Mike Baird to fund housing services based on data? Or on the number of men who kill women?
Baird has zero regard for evidence-based policy. This is a claim backed by evidence.
Similarly, merely counting the number of people in prisons and detention centres does absolutely nothing for the conditions for people in those places; and does absolutely nothing to decrease the rate of incarceration and detention (which are different things, according to the High Court of Australia; see Al-Kateb, critiqued here).
Trotting out some feel-good claim about more accurate measure of Aboriginal life expectancy and twinning this with generalised projections about data and better policy implies that the data collection will somehow improve morbidity rates for Aboriginal people.
But it won’t. The claims are disingenuous at best. Some would say such statements are grossly misleading. It is also harmful. This messaging is designed to create the impression that government cares about Aboriginal lives – when the opposite is true.
Some reflections on Homelessness an(d) the rough sleeper count
In 2011 I was employed by the Australian Bureau of Statistics as a Special Area Supervisor – Homelessness. The homelessness collection was an attempt to count as many rough sleepers as possible. There was no strategy to identify secondary homelessness. The ABS assumed that people who are couch-surfing would be counted in the households where they were staying that night.
This is naïve at best. A significant proportion of couch-surfers stay with friends or family whose rent is linked to the number of residents in the household. Those tenants are at risk of accumulating a debt, or even eviction, for extending a helping hand.
The causes of homelessness are complex, with multiple overlapping issues such as escaping violence, mental illness, and chronic addiction. But ultimately, homelessness can be traced to the breakdown of relationships, whether that is relationship with family, a landlord, an employer, or the state.
Couch surfing is tenuous and fraught, it raises a strong presumption that many relationships have already broken down.  Few people in this situation can afford to add more risk to relationships with a host or landlord or the state. To assume that vulnerable people are in a position to take on more risk for the sake of government data collection is to lack insight into their situation.
One of the worst types of homelessness is created by government: TA. Temporary accommodation is when the state pays for housing applicants to stay in a motel. It is very widespread, not a one-off stop-gap but a systemised response – and a miserable and expensive failure.
TA creates anxiety. People are compelled to spend every day applying for housing and must re-apply for TA eligibility every week. It creates gross discrimination. Many motel owners split their accommodation and provide vastly inferior services – plastic cutlery, no toilet paper – to TA guests, as though somehow government money buys less service for no reason other than exploitation of the poor – and the taxpayer.
It is a terrible dehumanising system that creates mini-ghettoes and resentment while motel owners net huge profits on the government purse. Motel owners who are of course small business and thus eligible for the $20,000 immediate write-down at our expense while touting their community as entrepreneurial and innovative and not at all like those lazy bludgers getting rich on Newstart.
No census can or will change this entrenched inequality, whether people in TA are counted or not.
For the 2011 rough sleepers count the ABS consulted widely and recruited the supervisors from the social housing sector.  The training was better than the consultation but both were predictably paternalistic. It seems it is virtually impossible for white middle class professionals to not reproduce unrealistic and ill-informed assumptions and stereotypes about the population with whom they work – yet it is the homeless who keep them in a job and their mortgages paid.
Some of those assumptions were around trust in authority. It is a familiar line – homeless people, or Aboriginal people, or young people, lack trust in authority. This is presented as some kind of deficit in the individual member of this or that community.
But the most cursory glance at the facts reveals that people who do not trust authority have reached an evidence-based conclusion: authority has, does, and will treat them badly. Violently. Brutally. Ignore their human rights. Deny their humanity. Be condescending and paternalistic and judgmental.
All of these are horrible experiences, and all are meted out, often, by government employees such as police; or government-funded employees, such as job agency staff.
So a mistrust of authority is a product of authority being oppressive; yet in the great Australian tradition this mistrust is framed as a deficit in members of the community to which government, historically and contemporaneously, has actively caused harm – usually under the guise of providing help.
The ABS had developed a two-tiered message for the homelessness count. The first was ‘trust us, we are trustworthy’. The second was ‘the data will assist government to make better policy which will benefit the homeless population’.
Fast forward to 2016 and the rules have changed, but the message is that same as that which informed our training for the homelessness count.
Using the same message for a different set of circumstances is lazy and complacent at best. At worst, it reeks of misleading the public: if there is a case to be made for the changes, why not make it? Why fall back on exactly the same message designed to engender a trust relationship with homeless people five years earlier?
Like all good tweeps, I put out a twitter poll: do you trust the government? There were two yes votes (n = 254). One person tweeted me to say she accidentally tapped yes.  Either way the yes vote was basically a margin of error. (No, I am not going to insult readers by spelling out the unscientific nature of a Twitter poll.)
Ironically, the failure to make the case for retaining names and addresses for a longer period is eroding trust in the ABS because the argument is so weak and the government so mistrusted.
The ‘better policy’ argument is specious for political expediency reasons already mentioned. The gap between the data collection and analysis and actual policy decisions, which are based on neoliberal ideology and electoral chancing, is huge. In addition, people most likely to not be counted are the people most likely to need government services to survive. Not government money in the form of research grants and public service jobs and immediate tax write-downs and public housing guests, but actual resources to feed and clothe and house themselves in a wealthy society.
Yet the line about better directing government policy based on census data is widely accepted… by people who do not rely on government services. The academics and bureaucrats pushing this line are not grounding it in evidence of better homelessness services, or identifiable improvements in the lives of welfare recipients. They are not doing this because the evidence is not there.
So like the trust argument acting to erode trust, the evidence-for-better-policy argument fails to point to evidence of better policy outcomes derived from the previous Census.
Meanwhile the evidence of a benefit to the academic or bureaucrat is there for all to see: there they are on the telly, with their job and media platform, well remunerated and recognised, as an expert in data capture and analysis.
Leaving aside identifiable groups of academics and bureaucrats, is the Census beneficial to homeless people? Unemployed people? Sole parents, carers, people with disabilities?
Are any of these groups better off than in 2011 because their status was counted and analysed by the ABS and other social scientists?
Why not ask them? I could easily find people who were counted as homeless in 2011. Pay me and I’ll let you know if any are better off, and if so how many. The government has moved many Centrelink recipients on to cashless welfare this year. Why not ask Alan Tudge if this policy is linked to census data? Ask people on the Basics card what they think of the alleged link between their Census form and having access to only 20% of their payment in cash?



No comments: