The legislation for the “cashless welfare card” trials passed the Senate last week with bipartisan support. The card started life as Twiggy Forrest’s proposal to control the spending of people on welfare payments under the assumption that recipients are not able to manage their finances. It will quarantine 80% of their payments and restrict users from spending anything on gambling and alcohol.
On the surface the bill looks reasonable. It establishes a 12-month trial program that would be evaluated. There is agreement that the communities targeted – such as Ceduna in South Australia – have serious issues that need attention. But it is unclear whether the solution is the cashless welfare card, or if there are wider issues.
Therefore, the trials need to be able to show the benefits or otherwise of the card. And the failure of a similar BasicsCard in the Northern Territory to produce clear benefits raises the need for careful design.
What will the trial actually achieve?
A 12-month trial of the card needs to offer the clear capacity to trace cause and effect. However, the trial will not follow this path, because it also includes funds for the same trial sites to receive improved local services.
This additional funding has been included in response to the Ceduna community’s conditions for giving its consent for the trial. Its submission to a Senate inquiry identified local service needs as core to any improvements in rates of alcohol abuse.
This agreement for including services locally creates serious issues for the trial’s claimed intentions. How can any form of research determine whether any improvements that may occur over 12 months are the result of the card or of the services, or a mix of both?
If this legislation is to fund a proper trial of a major change for delivering payments, one that may be used for much wider reform of welfare payments, this one will not be able to do so. Such a trial needs a design that will clearly distinguish the effectiveness of the card on its own.
Causal connections are always a problem for reporting on outcomes. These doubts are reinforced by excessive claims quoted in the Senate report on the bill. For example, Parliamentary Secretary to the Prime Minister Alan Tudge noted that:
… there will be a detailed evaluation process which will be undertaken. It will be an independent evaluation, and by and large we will be tracking the main harm indicators in the community as well as taking some qualitative assessments.
And in its submission to the Senate inquiry, the Department of Social Serivces said:
The trial is limited in scope – a trial in the true sense – and will include an independent comprehensive evaluation considering the impact of limiting the amount of welfare funds on community-level harm. The evaluation will include qualitative and quantitative data analysis providing clear findings for government and the communities.
The above claims make no sense. On the one hand, the government says this will be a rigorously evaluated 12-month trial of the card. On the other hand, it is promising substantially increased services to deal with the same problem of alcohol addiction to gain Indigenous community consent. How do you separate the card’s effects from the benefits from the services?
Given that some 800-plus recipients of welfare payments will involuntarily be covered in Ceduna and surrounding areas – and maybe only 60 to 80 have drinking problems – there will be other issues to be covered. What effects does the trial have on those who lose some control over how they spend money and where? Some evidence from the NT suggests that the process may cause damage to self-esteem.
This was also raised in the Senate inquiry. The Australian Association of Social Workers submitted that:
Within the trial sites there will be a large number of welfare recipients who manage their scarce resources well and who do not have a problem with alcohol, illegal drugs or gambling. Their normal patterns of financial management will be disrupted yet they will gain nothing from the trial.
Not learning from history
The extra services being funded as part of the deal are probably what is really needed, but it is unclear whether they will be funded after the trial. This suggests that the money involved should be allocated for services, as that is what the local communities believe is needed, before any tinkering with incomes.
No evidence was offered to the Senate inquiry that the welfare card’s predecessor – the BasicsCard – had produced any benefits for the acknowledged serious issues of alcohol abuse and violence in the Northern Territory. The official evaluation of income management:
… could not find any substantive evidence of the program having significant changes relative to its key policy objectives, including changing people’s behaviours … The evaluation data does not provide evidence of income management having improved the outcomes that it was intending to have an impact upon.
In the NT evaluations, a serious gap emerged between the results of some of the interviews with participants and the official statistics. The latter showed no attributable improvements in security, school attendance, health outcomes or other wellbeing indicators of those who had been subject to income management.
The existence of other changes as part of the Northern Territory Intervention also affected the results. The conclusions quoted above carefully decided that there was no reliable evidence that income management had been effective.