Evaluation of the Australian Wage Subsidy Special Youth ...

Evaluation of the Australian Wage Subsidy Special Youth ... Evaluation of the Australian Wage Subsidy Special Youth ...

02.06.2014 Views

216 10 Treated propensity scores, Untreated propensity scores, 5 0 0 .2 .4 .6 attrition weight kernel densities of weightd propensity scores, treated vs untreated Figure 6.6 Kernel density plot of attrition weighted propensity scores, before matching Note: Epanechnikov kernel.

217 Table 6.7 Matching results, single nearest neighbour with replacement, within caliper, weighting the propensity with combined weights for attrition, non-response and design Match with caliper width 0.001 Match with caliper width 0.005 Match with caliper width 0.01 Match with caliper width 0.02 Match with caliper width 0.05 Difference in employment 150 for 0.08 0.06 0.02 0.02 0.02 matched treated and comparisons T statistic 1.57 151 1.19 0.14 0.14 0.14 Number of SYETP matched 87 99 102 102 103 Number of comparison which satisfy 80 89 91 91 92 the caliper rule Number of times used 1 73 80 81 81 82 2 7 8 9 9 9 More than 2 1 1 1 1 Mean difference in propensity score 0.0002 0.0004 0.0005 0.0005 0.0009 between single nearest neighbour matched treated and comparisons Standard deviation 0.0002 0.0006 0.0012 0.0012 0.0035 Mean bias 11.04 11.85 11.73 11.73 11.84 Common support – 1 case dropped. Weighting protocol: weight propensity, weight the match using the treated weight only. The weighted mean bias is calculated using svymean in Stata. 150 Ever employed in 1986 survey. 151 Probability of (0.22) for acceptance of the null hypothesis.

217<br />

Table 6.7 Matching results, single nearest neighbour with replacement, within caliper,<br />

weighting <strong>the</strong> propensity with combined weights for attrition, non-response and design<br />

Match<br />

with<br />

caliper<br />

width<br />

0.001<br />

Match<br />

with<br />

caliper<br />

width<br />

0.005<br />

Match<br />

with<br />

caliper<br />

width<br />

0.01<br />

Match<br />

with<br />

caliper<br />

width<br />

0.02<br />

Match<br />

with<br />

caliper<br />

width<br />

0.05<br />

Difference in employment 150 for 0.08 0.06 0.02 0.02 0.02<br />

matched treated and comparisons<br />

T statistic 1.57 151 1.19 0.14 0.14 0.14<br />

Number <strong>of</strong> SYETP matched 87 99 102 102 103<br />

Number <strong>of</strong> comparison which satisfy 80 89 91 91 92<br />

<strong>the</strong> caliper rule<br />

Number <strong>of</strong> times used<br />

1 73 80 81 81 82<br />

2 7 8 9 9 9<br />

More than 2 1 1 1 1<br />

Mean difference in propensity score 0.0002 0.0004 0.0005 0.0005 0.0009<br />

between single nearest neighbour<br />

matched treated and comparisons<br />

Standard deviation 0.0002 0.0006 0.0012 0.0012 0.0035<br />

Mean bias 11.04 11.85 11.73 11.73 11.84<br />

Common support – 1 case dropped. Weighting protocol: weight propensity, weight <strong>the</strong> match using <strong>the</strong><br />

treated weight only. The weighted mean bias is calculated using svymean in Stata.<br />

150 Ever employed in 1986 survey.<br />

151 Probability <strong>of</strong> (0.22) for acceptance <strong>of</strong> <strong>the</strong> null hypo<strong>the</strong>sis.

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

Saved successfully!

Ooh no, something went wrong!