Working Paper

Do Welfare Restrictions Improve Child Health? Estimating the Causal Impact of Income Management in the Northern Territory

Published: 2017

Non-Technical Summary:

In 2007, the Australian government changed the way that welfare recipients in Indigenous communities in the Northern Territory were paid. Instead of receiving unconditional cash payments, recipients had half of their payments quarantined into a separate account which could only be used for priority goods. One of the stated goals of income management was to improve child outcomes by increasing the share of household income spent on food and other household essentials, and reducing the amount spent on potentially harmful goods such as alcohol and tobacco.

Over the past decade, governments, commentators and journalists have questioned whether income management has achieved this goal. Despite widespread use of such restrictions, and a large literature evaluating the impact of individual transfer programs, there is relatively little empirical evidence that directly compares the impact of restricted with unrestricted transfers. In this paper, we provide a first estimate of the impact of income management on child health using data on birth outcomes.

The findings of our study suggest that income management did not improve one measure of child health outcomes, and, by extension, that income management does not appear to have produced the desired change in household consumption patterns, at least for households with pregnant women. In fact, income management may have had a negative impact on newborn health – lower average birthweights and a higher probability of low birthweight (defined as less than 2500g), over and above what would be expected if a baby was premature.

Other research suggests that policy-driven changes in pregnant women’s nutrition have the largest impact on birthweight if introduced in the third trimester. But exposure to income management mattered most for birth outcomes when it occurred in the first or second trimester of the pregnancy, which suggests that channels other than lack of nutrition may have caused the adverse effects. We are unable to identify the channel of the treatment effect, but one possibility could be increasing levels of stress experienced by the mother when income management was implemented. The restriction may have disrupted existing intra-household financial arrangements, by changing the value of resources over which individual household members had discretionary decision-making power. This may have negatively affected consumption patterns or even caused conflict within the household. Another source of stress could have originated from policy implementation problems, such as confusion over how to access funds, the administrative and time burden of allocating funds to various uses, and general dissatisfaction with this policy as part of the NTER policy package (see Cobb-Clark, Kettlewell et al., 2017 who find similar negative effects on school attendance).

Our paper contributes to the literature in three important ways. First, it adds to the emerging international literature on the effectiveness of restricted transfers relative to cash transfers, especially for highly vulnerable and marginalised populations. Second, our research contributes to a wide literature that studies the long-term drivers and consequences of early-life health problems, especially low birthweight. Third, this research contributes directly to the debate on the effectiveness of income management in reducing health disparities.


Mary-Alice Doyle

Centre Member

Stefanie Schurer
Sven Silburn