Working Paper 34
Evaluating the outcomes of labour market programs (Part A)
Abstract
This paper introduces some statistical techniques for measuring the impact of labour market programs. A major complication to the impact evaluation of labour market programs is that some characteristics of participants of the program are different from those of non-participants. These characteristics not only influence the decision whether to participate in the program but also affect the outcomes of the participants and non-participants. Based on the potential outcomes framework, several parameters of interest can be unambiguously defined and the problems of comparing observed outcomes can be easily identified. Under the assumption that selection to the program is based on observable characteristics, 'matching' is shown to be a superior estimator of treatment effects than linear regression. A proposal to evaluate the Community Jobs Plan - Employment Assistance (CJP-EA) and Community Jobs Plan - Work Placements (CJP-WP) programs using matching methods is outlined in this paper. The forthcoming Part B of this paper will focus on the statistical techniques based on the assumption that selection to the program is based on characteristics which are unobserved by the evaluator.
Download Evaluating the outcomes of labour market programs (Part A) [PDF 178kb]
Contact Apprenticeships info 1800 210 210 or Training Queensland 1300 369 935
Other languages