Policy makers will be in a better position to facilitate the transition from welfare to work if they have a detailed understanding of the incentives facing the potential Indigenous workforce. Distinguishing the expected gains from becoming employed (captured by the replacement rate) from the expected loss of becoming unemployed (estimated in cost of job loss) is a prerequisite for developing a policy framework which ensures that more Indigenous workers have an incentive to look for, secure and keep jobs. This paper uses the 1994 National Aboriginal and Torres Strait Islander Survey data to analyse the issues involved.

Incentives for Indigenous Australians to work

The major work incentive issues raised by the literature are the effect of the welfare system on search activity, the duration of unemployment, the take-up rates to entitlements and the effect on the labour supply decisions of partners. The first two issues are directly captured by the replacement rate and cost of job loss measures.

• Ninety-four per cent of the Indigenous unemployed aged 18-64 years were receiving government benefits. This high take-up rate suggests that any conclusions about the relationship between the benefit system and unemployment are likely to be applicable to the majority of unemployed Indigenous Australians.
• The labour supply decisions of Indigenous couples are highly correlated. Among unemployed males, 91 per cent had partners who were either unemployed or not in the labour force. Among unemployed females, 40 per cent had partners who were either unemployed or not in the labour force.

Indigenous replacement rates

Replacement rates provide one summary measure of the incentive to work in the presence of the social security system. In simple terms, it can be considered to capture the immediate gains from employment for potential members of the labour force. The closer a replacement rate is to one, the less incentive an individual has to work. Replacement rates greater than one mean that an individual has no monetary incentive to work as they can receive more income from remaining outside employment. However, if working in a low paid job today leads to substantial increases in future wages, then it is possible that a person chooses to work despite the immediate financial disincentive.

• Replacement rates for single males and females are highest for those not in the labour force but do not differ substantially between those in non-Community Development Employment Projects (CDEP) employment or unemployment.
• The average single Indigenous male could expect to receive from unemployment benefits 69 per cent of the average income of a single Indigenous male in any non-CDEP employment.
• Replacement rates were higher for females than males and higher for those who were married or in a de facto relationship. The first result reflects the lower earnings of females compared with males and the second, the fact that the welfare system explicitly recognises the additional costs of supporting a dependant spouse while employers do not.
• Among single males and females, about 20 per cent of individuals could expect a higher income from welfare than from employment in non-CDEP work (that is their replacement rate was greater than one). Among those in a partnered relationship, the share was higher, 30 per cent for males looking for work with a dependant partner and almost 80 per cent for females looking for work with a dependant partner.
• The estimated replacement rates rise with the number of dependants, as the welfare system pays additional benefits to larger families while employers do not explicitly adjust their wages and salaries to take account of family characteristics.
• Large differences in the replacement rates arise from changes in the assumptions of how the alternative wage is calculated. This illustrates that the incentive to work may be very sensitive to the type of work available in the local labour market.
• Incentives to work do not vary much over the lifecycle for Indigenous males. Even though replacement rates are higher for older females, the results may be very sensitive to the assumptions made about the alternative employment available.

Cost of job loss among the Indigenous employed

The cost of job loss is estimated for people who are currently employed. This measure includes both the effects of duration of unemployment and the replacement rate on the costs of unemployment to the individual. Our calculations show that the costs of unemployment are particularly high for many Indigenous Australians because they tend to have longer durations of unemployment.

• Previous research concludes that the cost of job loss has gone up for all Australians over time but our calculations show that they were substantially higher for the vast majority of the Indigenous population than for Australians in general in the early 1990s.
• The expected cost of job loss was particularly high, relative to the Australian average, for partnered Indigenous Australians because of their substantially higher probability of being unemployed. So for example, the expected cost of losing a job for a partnered Indigenous male with no children was an income equal to 90 per cent of the income from full-year full-time work. Among all Australian partnered males in employment, the low probability of becoming unemployed meant that the equivalent statistic was 98 per cent.

Distinguishing the gains of becoming employed from the cost of job loss

The main advantage of the methodology used here is that it allows us to disaggregate the different incentives to find and stay in employment. For example, while the monetary incentive to find work, as measured by replacement rates, is lower for Indigenous Australians, they have more to lose once they are employed because of the long spells between jobs.

• One possible implication of this result is that Indigenous workers may be more reluctant to voluntarily quit than other employees. The resulting reduction in labour force mobility may have adverse impacts on the suitability of job matches in the Indigenous workforce.

The level of change required in social security payments to provide an incentive for all the Indigenous workforce to find and keep a job is probably politically and socially unacceptable. While the family payment reforms of the late 1980s and early 1990s provided some relief to the so-called 'working poor' families, significant labour market disincentives persist for many Indigenous families.

ISBN: 0 7315 2594 9

ISSN:1036 1774