02.06.2014 Views

Evaluation of the Australian Wage Subsidy Special Youth ...

Evaluation of the Australian Wage Subsidy Special Youth ...

Evaluation of the Australian Wage Subsidy Special Youth ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

86<br />

still much lower for <strong>the</strong> most-disadvantaged, <strong>the</strong>y were still worse <strong>of</strong>f than <strong>the</strong> leastdisadvantaged<br />

after <strong>the</strong> programme. This arose because <strong>of</strong> <strong>the</strong> difference in employment<br />

chances for non-participant– i.e. <strong>the</strong> quasi-control group for <strong>the</strong> most-disadvantaged and<br />

least-disadvantaged had very different employment chances. Table 2.20 shows <strong>the</strong>ir<br />

results for SYETP. The results show a positive impact on full-time continuous<br />

employment for all groups after SYETP participation, however <strong>the</strong> size ranges from 5 to<br />

55 per cent.<br />

Table 2.20 Rao and Jones (1986) Estimated percent post-programme full-time<br />

continuous employment chances 1981-1983<br />

Least Disadvantaged<br />

Most Disadvantaged<br />

Completed<br />

programme<br />

Difference to<br />

quasi- control<br />

Completed<br />

programme<br />

Difference to<br />

quasi-control<br />

SYETP<br />

63.4 44.8 8.2 7<br />

Commonwealth<br />

SYETP private 72.0 53.4 11.7 10.5<br />

2nd SYETP 55.7 37.1 6.1 4.9<br />

Extended SYETP 74.4 55.8 13.0 11.8<br />

Quasi-control 18.6 1.2<br />

Source: Rao and Jones (1986) Table 3, p24<br />

This analysis made an effort to define comparison and treatment groups based on<br />

observed characteristics, while lacking true control data in a survey <strong>of</strong> participants. The<br />

quasi-control would be subject to a reasonably large extent <strong>of</strong> contamination bias, as <strong>the</strong><br />

data are based on survey data recalling labour market history. Contamination bias is<br />

where <strong>the</strong> treatment and comparison groups are not cleanly defined and mixing can occur.<br />

There is also <strong>the</strong> strong possibility <strong>of</strong> selection bias. The Baker (1984) results showed<br />

some evidence that completers and non-completers had different characteristics, and had<br />

different raw employment outcomes. Rao and Jones (1986) confirmed that controlling for<br />

individual characteristics was essential. It is clear that individual characteristics might<br />

also influence which programme an individual participated in however this was not<br />

controlled for. There was no modelling <strong>of</strong> programme entry. Again, <strong>the</strong> various<br />

programmes were simply dummies in <strong>the</strong> employment equation, which is subject to <strong>the</strong><br />

same difficulties as <strong>the</strong> Baker (1984) analysis. Baker (1984) showed that non-response

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!