PC News - May 2015

Housing assistance, delivered as public housing and Commonwealth Rent Assistance (CRA), has a broad reach. In 2013, about 700 000 Australians lived in public housing and a further 2.4 million who lived in private rental accommodation benefited from CRA.¹

For a range of reasons, it might be expected that housing assistance affects whether a recipient works.

• For most public housing tenants, rents are set as a proportion of income, so a tenant's rent rises when their income increases — a potential disincentive to employment.
• In contrast, CRA recipients pay market rents and CRA isn't withdrawn until a recipient is earning a relatively high income. CRA increases a recipient's real income, so might influence work decisions.
• Lease terms are typically longer in public housing than in the private sector and the associated housing stability might mean that a public housing tenant is better able to work or study than would be the case in other tenures. The stress and uncertainty associated with a lack of stable housing are likely to reduce a person's ability to search for work, for example.
• Some public housing and lower cost rentals might be located in areas with poor access to transport or jobs and a concentration of disadvantaged people. Employment outcomes might be affected, for example, if high rates of social exclusion in a neighbourhood mean that residents don't have strong family and social networks to support them in accessing employment opportunities.
• Rules about entry to public housing might affect the work activity of people on waiting lists. People can typically only enter public housing if their income is low. Applicants for public housing may choose not to work or work fewer hours than otherwise, so that their income remains low enough for them to qualify for a property. Similar disincentives could apply to tenants if continued tenure is contingent on remaining below an income eligibility threshold.

A recent Commission Research Paper has drawn on three large confidentialised administrative datasets to examine the links between housing assistance and participation in employment. Access to Centrelink administrative records on all income support payment (ISP) recipients for the period 2003-13 enabled a more detailed look at the relationship between housing assistance and employment than has been possible previously. Access to administrative records for public housing applicants and tenants in South Australia and Western Australia has enabled research into how moving into public housing might affect the employment of people who became tenants over the decade to 2013 — a period that saw increasing priority for high needs applicants in housing allocations.

A quick look at the Centrelink data for ISP recipients suggests that housing assistance, or at least public housing, might be related to participation in employment. At 30 June 2013 about 20 per cent of all ISP recipients worked, but the employment rate of those living in public housing was only half that level — 9.8 per cent. Lower rates of employment among public housing tenants relative to CRA recipients and other income support recipients who don't receive housing assistance are seen across all ISP types (figure 1).

But marked differences in employment rates between public housing tenants and other ISP recipients remain, even when these observed characteristics are taken into account.

It is also possible that public housing tenants are different from other ISP recipients in ways that are not easily observed in the data, and that these differences matter for employment. Perhaps public housing tenants are more likely to have drug or alcohol problems, a criminal record or mental health issues, for example. Public housing providers have increasingly prioritised more highly disadvantaged people — including those who are homeless or at risk of homelessness — in housing allocations.

To the extent that these unobserved characteristics don't change over the time period studied (or are strongly associated with characteristics that don't change over the period), statistical techniques can be used to account for their effect on employment — even if the nature of the characteristics is unknown. When this is done, differences in employment rates between ISP recipients who do and do not receive housing assistance become very small (figure 2).

In other words, the relatively low employment rates recorded for public housing tenants are due to their characteristics, not their receipt of housing assistance.

• Figure 1 Employment of housing assistance recipients by income support payment type^abc
  Per cent reporting earned income

  
  1. ISP recipients aged 15-65.
  2. Other payments include a range of less common income support payments, including Bereavement Allowance, Wife's Pension, Wife's Disability Support Pension, Austudy, Partner Allowance, Sickness Allowance, Special Benefits, Widow Allowance and Abstudy.
  3. Employment rates are based on a pooled sample spanning from 2005 to 2013. These figures are not directly comparable to those in table 1 where the focus is on working age ISP recipients active at 30 June 2013.
  Source: Commission estimates based on Research and Evaluation Database.

• Figure 2 Expected effect of housing assistance on employment, by Income Support Payment type^a,b
  Percentage point difference from no housing assistance, 2005-2013 data, accounting for observed and unobserved characteristics

  
  1. The employment effect of housing assistance is calculated using odds ratios from a fixed effects logit model that includes interaction terms between housing assistance type and ISP, and takes into account unobserved differences between ISP recipients.
  2. Other payments include a range of less common income support payments, including Bereavement Allowance, Wife's Pension, Wife's Disability Support Pension, Austudy, Partner Allowance, Sickness Allowance, Special Benefits, Widow Allowance and Abstudy.
  Source: Commission estimates based on Research and Evaluation Database.

Despite the availability of employment assistance programs, the employment rates of public housing tenants have changed little over the past decade. Existing programs appear not to be particularly effective at addressing the characteristics that are impeding employment for many public housing tenants.

A new approach may be needed to tackle the low levels of employment among ISP recipients living in public housing. Such an approach could combine intensive support for tenants looking for employment with temporary changes to the often high effective marginal tax rates that they face when entering the workforce.

Among other findings, the research also revealed:

• A strong link between housing stability and participation in employment. Taking other characteristics into account, ISP recipients who move twice over the course of a year, for example, are predicted to have an employment rate nearly 6 percentage points below that of their peers who do not move — a large difference in the context of an employment rate for non-movers of about 21 per cent.
• That public housing tenants who live in highly disadvantaged neighbourhoods are less likely to be working than those who live in less disadvantaged neighbourhoods, but only slightly — at 0.6 percentage points the effect is quite small.
• A rise in the employment rates of successful applicants for public housing in both South Australia and Western Australia while they were on the waiting list — suggesting that the potential problem of applicants avoiding employment while waiting for public housing in order to remain eligible should not be a major concern in these states.