Effects of activity test arrangements on exit from payments: the 9-month intensive review

Abstract

Since 1996, recipients of unemployment-related welfare payments in Australia have been subject to a review process when their payment spell duration reaches 9 months. This review process is both a monitoring and counselling device: payment recipients are required to provide details of job search activity, while the payments administrator (Centrelink) provides job search advice and assistance. Using Centrelink administrative data over the period 1995 to 2000, this study examines the effects of these reviews on exit from unemployment-related payments. Limitations of the administrative data – in particular, the absence of information on the review process and the nature and precise timing of the review for each recipient – constrain the choice of empirical method. We therefore use duration analysis methods – specifically, estimation of empirical hazards and hazard models – to indirectly infer the impact of the review. Two alternative empirical approaches are taken. The first compares the hazard rate at the 9-month spell duration with hazard rates at 'nearby' spell durations. The second approach exploits a policy change which occurred in March 1996, when the review timing changed from 12 months to 9 months spell duration. For this approach, we compare the 9-month hazard rate in the post-March 1996 period with the 9-month hazard rate in the pre-March 1996 period. We do not find compelling evidence of a substantial or significant effect of the 9-month review using either of the empirical approaches.