The economic evaluation of Youth Training Programmes

• Main Author: Fargher, Scott W.
• Published: University of Edinburgh 2002
• Subjects: 331.2
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.650495

This thesis investigates issues associated with the economic evaluation of government intervention in the youth labour market. Specifically, it focuses on the evaluation of youth training programmes and assesses the sensitivity of post-programme impact estimates to methodological choices embodied in different non-experimental methods. Relatively high and persistent youth unemployment rates, together with rapidly changing patterns in the demand for skills, demographic change and social exclusion fears, have led to the expansion of government provided education and training opportunities in most modern economies. Moreover, the increased application of ‘active labour market policies’ combined with the desire to maintain international competitiveness has contributed to a greater emphasis on training in most industrial countries. Clearly, the outcome of evaluation research has an important bearing on the course and scale of such intervention and is necessarily an integral part of any active labour market policy. The use of non-experimental techniques in evaluation research has been widely questioned in light of divergent cross-study impact estimates, even when applied to seemingly similar programmes and data together with inconsistencies when compared to estimates derived using experimental techniques and data. This thesis investigates different non-experimental methods used in economic evaluation through the detailed examination of government assisted youth training in Scotland supplemented by an examination of training provision in Australia. The experimental/non-experimental debate is used as a framework to review recent theoretical developments in the evaluation field. The analysis begins with an examination of issues in the youth labour market that have lead to increased levels of intervention. Different evaluation methodologies are then outlined followed by a detailed exploration of the institutional developments in the youth labour markets of both the United Kingdom and Australia. This is complemented by a critical review of the relevant empirical literature focusing on potential sources of cross-study variation in reported impact estimates. A meta-analysis carried out on evaluation studies in the United Kingdom confirms that a substantive part of the variation may be due to the model employed. To investigate this issue in more detail the thesis then employs sensitivity analysis to determine the impact of model choice on a common database. The empirical analysis draws on two distinct longitudinal datasets the Scottish Young Person’s Survey and the Australian Youth Survey and examines initial returns to training in the early 1990s.