Income management is coming to a place near you — eventually. The whole Northern Territory is already an “experiment” as more welfare payment recipients (indigenous and not) are being included in the program, willy nilly.  The extension of the program has occurred despite strong opposition from many sources and a singular lack of valid evidence of improvements in child well-being in the 72 areas already covered. The minister, Jenny Macklin,  yesterday also announced that NT parents with child problems may also be income managed. To quote:

A new tool to help ensure parents are providing their children with the necessities of life is now available across the Northern Territory. As part of the Australian Government’s roll-out of non-discriminatory income management across the Territory, we have introduced child protection income management.  This means Territory Government child protection workers can now refer parents to Centrelink for compulsory income management when children are being neglected or are at risk of neglect.

There is already a similar “pilot” program for child protection cases already operating in Western Australian, alongside a voluntary IM program. An evaluation of these was also released yesterday with the following media release comments by the minister.

An independent evaluation has shown the Australian Government’s trial of child protection and voluntary income management in Perth and the Kimberley is having a positive impact on the well-being of children and families in Western Australia …

The evaluation shows income management is helping to improve the lives of families in Western Australia by ensuring welfare is spent where it is intended — on the essentials of life and in the best interests of children.

The evaluation of the WA income maintenance programs offers some interesting indications of potential benefits and limits of IM. The two programs are very different and are also different from the Cape York version or the NT one. There are serious limits, however, to using the data for generalising and assuming clear benefits.

First, the numbers are limited, so standard errors are high (12%+ as there are only 149 interviews across three samples and two sites.

Total
(interview component)
Universe Sample
completed
Number of
people who participated in the survey as a percentage of all people in
the sample who were contacted.
Perth 595 74 72 83%
Kimberley 1145 84 77 79%
All 1740 158 149 81%

Secondly, there is no clear outcome data, because all the statistics offered are based on opinions of participants or service deliverers. No data from store records, medical examinations, case files or bank accounts are included. The responses are to precoded questions, which, given they know the purpose of the survey, may be geared to interviewer expectations.

Respondents were also paid $50 so responses again may be skewed to pleasing interviewers.

The so-called focus group stage was apparently highly structured and numbers are too limited to be useful as more than a vague indicator. The results are therefore to be treated as interesting but not reliable enough to be seen as proof the program is improving children’s lives. Differences in responses from the “control group’ may be sometimes statistically significant but are sufficiently slight to be no more than interesting. As the control group also reports some improvements in money management, questions of what other interventions may have had some effect also exist.

This suggests unmeasured factors such as the impact of the available financial counselling, budgeting and other supports that are part of the package. Would these services alone, and maybe other programs they may have had on nutrition, have had similar affects without income management?

The interviews with staff involved in delivering the program show their firm convictions that the program is working but this is not reliable because they have a vested interest in claiming that their clients are benefiting. This is where material should have been collected from case records e.g. recorded actual changes in behaviour, or financial records and store spending data would be necessary to validate their potentially self-interested claims.

And despite all the advantages the methodology seems to offer the client, FaHCSIA, the results show relatively small improvements and some negatives as well. The best the independent consultants could seriously claim from this research is that, when asked, most of the compulsory and voluntary participants felt they had gained something from the program, and some, but not all ex-IM participants claimed they retained some extra skills.

The costs of the program are not put into the report, so there is no way to assess cost benefits.  However, it is a program that offers more personal involvement of staff than the NT compulsory universal version. The latter is assumed to cost about $80 per week per person to administer, and this would presumably cost more, both because it is selective and includes financial support staff and child-protection liaison for some clients.

Expenditure on staffing rather than on clients directly seems to be a major problem in these areas. Yesterday’s Australian claimed “In the three years since the Howard government launched the intervention into 73 remote NT communities, the ranks of the NT-based federal bureaucracy have swelled by 180. There has been a combined increase since June 2007 of 609 staff of FaHCSIA based in the NT and public servants employed in the NT Department of Local Government, Housing and Regional Services.”

These numbers indicate that the government is more ready to fund bureaucrats than welfare recipients. These still need extra money to meet bills (and the data shows they often are lacking this despite new budget skills). So why not do a cost-benefit study on the costs of quarantining income voluntarily or compulsorily versus the benefits of more cash, direct services for children  and some widely available financial education for the same groups?

Assuming that the recipients of welfare payments need and benefit from income management  is not clearly shown, either in this study or any other done by FaHCSIA so far.