The cost of homelessness and the net benefit of homelessness programs: a national study

The results support the contention that persons at risk of homelessness are heavier users of non-homelessness services than the population in general. The potential annual cost offset per client if health, justice, welfare, children in care and eviction rates were to be reduced to population averages, ranges from $14 712 per client/year for street-to-home services to $44 137 per client/year for single men. If this offset were able to be maintained over the average remaining lifetime, this equates to a cost offset of between $352 826/client for street-to-home and $1 058 491/client for single men. Even if cost savings only relate to a five-year period, they range from $67 377 for street-to-home clients to $202 135 for clients of services for single men.

 

The largest cost to government comes from use of health and justice services and welfare benefits. The cost of children being placed in care is also large for clients of single women’s and tenancy support services. The cost of eviction from public housing is not a large cost per client. However, this is because, except for tenancy support programs, only a small proportion of clients have been in a public housing tenancy. Where a person had a public tenancy the incidence of eviction and associated cost is high. In all programs hospital stays represent one of the largest drivers of the cost differential, being 52 per cent of the total difference in health care costs, and thus one of the largest potential cost offsets if utilisation is able to be reduced to population average. The pattern in justice offsets is different for single men compared with the rest of the programs examined. Clients of single men’s services exhibit a much higher incidence of being held in prison, remand or detention, accounting for 66 per cent of the difference in justice costs. For all other programs the highest cost differentials are for ‘Victim of assault of robbery’ and ‘In Court’, accounting for 80 per cent or more of the higher than population average justice costs observed for these client groups. The unemployment rate for persons available to work is between 75 and 100 per cent for all programs, resulting in a cost to government from Newstart payments and lost tax receipts across all case managed clients of $15 923/year per person available to work, or $6620/year per client. The large difference here relates to the low 42 per cent labour force participation rate of clients, where a higher than population average are eligible for benefits such as DSP, and whose income source is considered unlikely to change with accommodation circumstances.

 

Indicative evidence is found that a link may exist between health and justice service utilisation and Indigenous status; with Indigenous respondents generally reporting a lower use of health services than non-Indigenous respondents, but a higher rate of contact with justice services.

 

This is particularly relevant for clients of single men’s services, where the high health cost of single men appears to be driven by non-Indigenous men, and the high cost of justice services incurred by this cohort appears to be driven by Indigenous men. However, the size of the Indigenous sample is very small and further research into this issue is warranted.

 

The factors that lead to homelessness are complex and, on average, the characteristics of persons at risk of homelessness differ from the population in general. Therefore, offsets estimated by comparing non-homelessness service utilisation by respondents with that of the population are likely to over-estimate achievable offsets. To address this issue, the health and justice service utilisation of persons who had experienced homelessness in the previous year are compared with those who had not. Again, outcomes for single men differed from other programs. Single men who had not experienced homelessness actually incurred higher costs than those who had not experienced homelessness. However, these higher costs for men who had not experienced homelessness related to circumstances where they were accommodated in high-cost institutional settings: hospital, remand or detention or in prison. Thus it cannot be argued that the cost to government is less for single men who experienced homelessness. In contrast, clients of single women’s and tenancy support services who had experienced homelessness display higher total costs of health and justice services than those who had not experienced homelessness.

 

Although point estimates of cost offsets must be treated with care, the pattern in health and justice offsets is largely consistent with those found in the WA study conducted by Flatau et al. (2008)25; in particular, the heavy use of high-cost hospital services by all groups and the different pattern in justice service costs observed for clients of single men’s services compared with single women and tenancy support. The findings of heavy use of high-cost hospital services by single men along with high rates of detention, imprisonment and time in court are also consistent with those from the Michael Project (Flatau et al. 2012).