Go to top of page

Centrelink's robo-debacle is a litany of inhuman errors

4 January 2017
Peter Martin

One of their stupidest mistakes is to calculate fortnightly income by dividing annual income by 26. If the figure is too high the robots say someone wasn't entitled to benefits during the weeks they received them, even if during those weeks the person earned nothing. In other words, they misapply the law. Another is that they are not too bright. If the name of an employer is spelt one way by the Tax Office and another way by Centrelink, the robots assume it's a different employer and that it's undeclared income. In other words, they shouldn't have been let loose.

How they came to be let loose, how they were allowed to shake down vulnerable people in the lead-up to Christmas, will doubtless be the subject of a Senate inquiry and probably an Audit Office inquiry.

There were clues on the Tuesday before the election. That's when Porter and Treasurer Scott Morrison said they had found billions to pay for their promises. Through "the smarter use of technology" they were going to "improve the capability for the identification and recovery of debt owed to taxpayers".

Automated compliance systems would "minimise red tape, and avoid mistakes that may adversely affect a recipient's payments".


One of the wilder theories is that they intended to. By inflicting a faulty debt recovery system on the public, they wanted to persuade ignorant, scared and busy people to hand over money they didn't owe and dissuade others from ever applying for benefits again.