Where to for income management?

Continuing this week's theme of social security and economic participation, Don Arthur asks about the future of income management within the Government's welfare strategy. This post originally appeared on the Parliamentary Library's FlagPost blog.

Patrick McClure’s February 2015 report on welfare reform urged a cautious approach to the expansion of income management, arguing that it should be used ‘judiciously’ and recommending that any changes should be informed by evaluations.

Income management is a scheme that sets aside a proportion of a person’s income support payment to stop them spending it on alcohol, tobacco, gambling or other excluded goods and services. It is intended to make sure people are able to meet their ‘priority needs’ and the needs of their families.

Income management doesn’t apply to everyone on income support. Firstly, it only applies to people who live in certain areas of Australia. Secondly, it only applies to particular individuals or particular categories of income support recipient in those areas.

Income management is applied through a range of measures. In addition to Voluntary Income Management, there are two kinds of compulsory measures:

• case by case measures where a decision maker considers a person’s individual circumstances before referring them for income management (e.g. Vulnerable Measure (social worker assessed), the Child Protection measure and the Supporting People at Risk measure)
• membership of a category measures where a person’s individual circumstances are not considered. A person is referred to income management based on criteria that can include the payment they receive, how long they’ve received it, and their age (Disengaged Youth measure and the Long-term Welfare Payment Recipients measure).

The evaluation research released so far suggests that income management is most effective when it is voluntary or is applied to participants after considering their individual circumstances. However experience shows that assessing clients individually is time consuming and costly and that relying on volunteers limits numbers. Both could prevent the scheme achieving economies of scale.

The Northern Territory (NT) has the largest number of people on income management of any location in Australia. With over 20,600 participants, the NT has more than four times the number of all the other income management sites combined. The NT also relies more heavily on targeting by category than other sites. However, according to the most recent evaluation of income management in the NT:

There was no evidence that targeting income management on the basis of duration in receipt of income support payment provides an effective basis for identifying those with particular vulnerabilities or a low level of money management skills. Similarly, there is no evidence that the range of income support payments at which Compulsory Income Management is targeted reflects the groups at highest risk. Compulsory Income Management is imposed upon a large group of people whom income management does not assist. This imposes costs upon those subject to income management and to the government (p. 319).

Consistent with the NT finding an August 2014 evaluation of the Place-based Income Management sites found evidence that Voluntary Income Management participants benefit from the scheme but failed to find benefits for clients who were referred to income management based on their membership of a category.

While the evaluation evidence released to date suggests that voluntary and case by case targeted income management is more effective than targeting based on membership of a category, the Government has proposed legislation which would mean that Centrelink social workers no longer undertake case by case assessments for the Vulnerable Measure of income management (however there will still be case by case assessments in Cape York and under income management measures such as Child Protection).

While the current per person cost of income management is high, that cost is likely to fall as the number of participants increases. According to officers from the Department of Social Services (DSS), a significant proportion of the cost of income management is tied up in fixed costs that don’t change as participant numbers increase or decrease (for example, the cost of setting up a 24 hour hotline for BasicsCard users) p. 82). The cost of each new participant is lower than the average cost because only the variable costs are affected (for example, the cost of additional staff hours for the hotline) (p. 72).DSS Secretary Finn Pratt has said that ‘If the government were to choose to expand income management substantially then the cost per person would go down radically’ (p. 77).

If the Government were to expand the program substantially, it is unlikely it would rely on volunteers or case by case assessments. One option would be to combine Centrelink’s administrative data with data from programs such as Emergency Relief to identify clients most likely to experience the kinds of problems income management is designed to solve. The New Zealand Government is taking an increasing interest in data and analytics including techniques such as predictive risk modelling. While this kind of approach could improve the cost effective of income management, it would almost certainly mean that that it would include more clients who may not helped by income management. In a recent inquiry, the Commonwealth Ombudsman raised concerns about automated decision-making processes in income management.

Two more evaluation reports on Place-based Income Management are due to be released this year (DSS was due to receive one report in January and the final report in May). The findings are likely to have implications for future decisions.

Don Arthur works in the social policy section of the Australian Parliamentary Library. This post originally appeared on the FlagPost blog.