*Epidemiologists everywhere scream* Seriously, this statistic by itself is totally meaninglesshttps://twitter.com/bugwannostra/status/1096539431874355200 …
-
Show this thread
-
Replying to @GidMK
Thank you! The source data is here, although it doesn't, as you point out, include the overall total — around 410,000
#robodebts — or cause of death. https://www.aph.gov.au/api/qon/downloadattachment?attachmentId=320c24fa-915f-40c1-8d8d-d816dd35f95b …pic.twitter.com/Q3bXwMbmFQ
2 replies 1 retweet 3 likes -
Replying to @not_my_debt
That's a more useful figure. So 2,000 deaths in a population of 410,000 over 3 years? That's very roughly equivalent to the Australian average, but without a lot of analysis hard to extrapolate further
2 replies 0 retweets 3 likes -
Replying to @GidMK
DHS don't go out of their way to be useful with figures
We'd want comparison data for non-debt Centrelink recipients to dig into the gender discrepancy, for instance, but that's still crude — current & former recipients aren't differentiated, & that may (or may not) matter.1 reply 1 retweet 2 likes -
Replying to @not_my_debt
Pretty much. You'd want a matched cohort of current centrelink recipients with age, gender and the like to even come cost to establishing causality. In my professional opinion, it's likely that there were some deaths caused directly by debt notices but probably not that many
1 reply 0 retweets 1 like -
Replying to @GidMK
We are in contact with family and friends of 5 people. But we shouldn't need a *death toll* for it to be imaginable that a system that uses a reverse onus + averaging to manufacture frequently inaccurate debts against vulnerable people might be… bad.
1 reply 4 retweets 4 likes
True, it's certainly not a great model, but I suspect for individual cases it will be impossible to prove causation. We might be able to demonstrate an increased risk of death with a statistical model, but too many factors in individuals
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.