What is the cost to government of homelessness programs?

Australian Journal of Social Issues, Winter, 2008 by Kaylene Zaretzky, Paul Flatau, Michelle Brady

1. Introduction

The net cost of providing support to clients of a social program is the gross cost net of any savings achieved elsewhere resulting from program participation. (2) Savings, or cost offsets, occur when social programs lower outlays in non-program related areas and/or increase revenues. In this paper, we provide indicative estimates of the net cost, including the opportunity cost of capital employed in providing accommodation, of providing a range of homelessness programs in Western Australia. Homelessness programs may improve the health, well-being, financial security, labour market and accommodation outcomes of clients. This, in turn, may result in decreased utilisation of homelessness prevention and support services in the future, reduced utilisation of hospital, justice and other services, reduced child residential care costs, lower housing authority and private rental management costs from the avoidance of evictions, lower income support payments and higher revenue from increased income tax payments. The analysis in this paper is focused solely on health and justice expenditure-based cost-offsets.

We base our findings on the net cost of homelessness programs on three sources. First, responses on the utilisation of health and justice services drawn from two surveys of clients of homelessness programs we conducted in Perth and the South-West of Western Australia. Second, information supplied by programs and agencies on the funding of homelessness programs and the cost of delivering services. Third, health and-justice unit cost data and general population-level health and justice utilisation estimates drawn from a range of different sources. We estimate that, on the basis of health and justice expenditure cost offsets alone, homelessness programs, in a given year, have the potential to save over twice the value of funding of homelessness programs. This estimate includes recurrent funding and the opportunity cost of capital employed in providing accommodation. The potential savings are much greater if changes in 'whole of life' service utilisation are considered.

Our estimates of the net cost of homelessness should be treated as indicative because data limitations prevent a fully comprehensive treatment of net costs at this point. The two key limitations are as follows.

First, we focus only on two types of cost offsets, namely, health and justice savings. We did not gather primary data and/or unit cost estimates on other potential savings (e.g., in terms of family-based homelessness, reduced placement of children in residential care). This results in an underestimation of the full value of cost offsets.

Second, our estimates of the net cost of homelessness assume that homelessness programs will reduce clients' use of health and justice services to a point where it is similar to that of the general population. In other words, the value of cost offsets are estimated as the difference between the health and justice expenditures for the homelessness program client population prior to their entry into the homeless programs, and the health and justice expenditures of the general population during the same period. This assumption is unlikely to be realistic because our survey results indicate that a significant proportion of clients entering homelessness programs have long histories of homelessness, high needs and poor quality of life outcomes. Because it is highly unlikely that following the provision of support, the average health and justice utilisation rates for these individuals will decrease to a level close to that of the general population our estimates potentially over-estimate the value of realisable cost offsets. (3)

However, we also estimate the proportion of potential cost offsets that would need to be realized for homelessness programs to be cost neutral. Only a small fraction of the potential health and justice offsets need to be realised for the benefits of providing homelessness programs to be greater than the cost. This fraction would be even smaller if the value of cost offsets other than health and justice cost offsets were included in the analysis. Our survey results also indicate that clients of homelessness programs gain important immediate benefits from homelessness prevention and support programs. (4)

Two major literature reviews of the cost and cost-effectiveness of homelessness prevention and support programs have been published recently (see Berry et al. 2003, and Pinkney and Ewing 2006). Both of these reviews point to a paucity of Australian evidence on the cost of delivering homelessness programs and, more particularly, the overall cost-effectiveness of homelessness programs. (5) This study aims to help fill this gap in the literature.

The structure of the paper is as follows. In section 2, we detail briefly the operation of the relevant homelessness programs in Western Australia and consider our survey findings on clients needs on entry to support and the short-term effectiveness of homelessness programs. Section 3 provides an outline of the study's method for estimating costs and cost offsets. Sections 4 and 5 provide a detailed examination of the potential cost offsets from the provision of support and net program costs, section 6 concludes.