Ombudsman slams Centrelink debt recovery system

Testing was inadequate, staff training worse, averaging incomes was known to be risky but used anyway

Australia's Commonwealth Ombudsman has published its report (PDF) on Centrelink’s automated debt raising and recovery system and found that while it is capable of correctly assessing debts, the agency was aware it had flaws and did little to prevent them.

The system was introduced by Australia's Department of Human Services (DHS) in July 2016 with the aim of recovering incorrect payments made under Australia's income support schemes. Those targeted under the scheme were directed to use an “online compliance intervention” (OCI) system to asses their debts. In early 2017 the system became the subject of considerable controversy as customers complained it was incorrectly calculating debts and that explaining that to Centrelink dumped them into a bureaucratic whirlpool and threatened them with penalties as they tried to escape by explaining the true nature of their debt (or lack thereof).

Many of those who hit trouble with the program did so because it used a small sample of data about their incomes and averaged it to assume annual incomes well beyond those that individuals received. Centrelink then calculated eligibility for payments on the basis of an annual average, rather than nearly-always-lower part-of-year-incomes that made its customers eligible for payments.

Averaging was done if Centrelink received inadequate data, or if it received no response to its communications with clients. Many non-responsive clients reported that Centrelink had tried to reach them at old addresses and therefore received debt notices based on averaged incomes.

Now the Ombudsman has found several deficiencies in the program. Acting commonwealth Ombudsman Richard Glenn's report does find that OCI is capable of calculating debts correctly. But he also finds that it was not easy to get the right data into the system because letters sent to Centrelink didn't explain income averaging.

Here's the Ombudsman's assessment of those letters:

Our investigation revealed DHS’ initial messaging to customers through its letters and in the system itself, was unclear and did not include crucial information such as a contact phone number for the DHS compliance team. Many complainants did not realise their income would be averaged across the employment period if they did not enter their income against each fortnight. This resulted in people with fluctuating or intermittent income having their income averaged. In some cases this was a more favourable outcome for the customer and in others, the debt was overstated.

The report also says that DHS had previously treated income averaging as a measure of last resort, because its own guidelines recognised it would likely generate over-estimations of debt. At section 2.33 the report offers this account of its investigation:

We asked DHS whether it had done modelling on how many debts were likely to be over-calculated as opposed to under-calculated. DHS advised no such modelling was done.

If every possible means of obtaining the actual income information has been attempted, it is possible to use any evidence available to raise a debt including an annual figure.

Those who copped an overstated debt then had to sort it out with staff who had not been properly trained in the nuances of the OCI program. Here's how the report explains that mess:

Poor service delivery was a recurring theme in many complaints received by our office. Customers had problems getting a clear explanation about the debt decision and the reasoning behind it. As the compliance helpline number was initially excluded from letters and was not obvious in the system, customers called general customer service lines resulting in long wait times. They could not always get clear information and assistance to use the online system. Service centre staff did not always have sufficient knowledge about how the OCI system works, highlighting a deficiency in DHS’ communication and training to staff. In some instances, a more thorough manual intervention by a compliance officer would have saved the customer time and effort.

Those messes, the report finds, were avoidable if OCI had been properly-planned and tested. The report says “In our view, many of the OCI’s implementation problems could have been mitigated through better project planning and risk management at the outset. This includes more rigorous user testing with customers and service delivery staff, a more incremental rollout, and better communication to staff and stakeholders.”

There's more – about 123 pages in all – but the thrust of the report is best-expressed by the recommendations it makes and there's eight of those that Vulture South summarises as follows:

1. Centrelink should not automatically impose a ten per cent debt recovery when pursuing OCI debts;
2. Initial letters sent under the OCI should put the help-line number front and centre, mention the possibility of a debt earlier to get readers' attention, clearly explain averaging and explain how to get an extension on response deadlines;
3. Explain averaging in the OCI system and “advise that debts based on averaged ATO income may be less accurate than debts based on actual income”;
4. DHS should do the heavy lifting to secure details of actual income, because it has powers to do so and will therefore lighten the load on clients and likeley produce more accurate income assessments;
5. Fix its helpline by training staff properly, ensuring there are enough staff to handle calls in a reasonable amount of time, offer clients proper guides on how to use OCI and make it plain to clients that they can get help on the phone;
6. Offer “staff-assisted interventions” to more clients;
7. Offer more assistance to vulnerable customers;
8. Evaluate OCI before expanding it and in future expand it only incrementally.

Overall, the tone of the report is strongly critical of the OCI program because it finds it was not well-designed and DHS' efforts to communicate it to clients and staff were both inadequate.

The report also finds that the program's flaws weren't taken seriously until they became public knowledge,

The report also finds “There will inevitably be problems with the rollout of a system of this scale,” adding that “In our view the risks could have been mitigated through better planning and risk management arrangements at the outset that involved customers and other external stakeholders in the design and testing phases.” And also through following DHS' own guidelines about income averaging.

But perhaps the scariest thing about the report's conclusions is how close they are to others on Australia's failed online census, which in an act of poor planning failed to properly consider an entirely-foreseeable set of circumstances and also failed to communicate well with the community. ®