Research topic: Program and policy evaluation

The program and policy evaluation and design research group was formed as a collaboration between the Melbourne Institute and the Department of Economics at the University of Melbourne in 2003. It aims to be a leading research centre on program and policy evaluation and design in Australia.

The area of program and policy evaluation and design involves application of experimental and 'quasi-experimental' empirical methodologies to evaluate outcomes from government programs and policies. These methodologies have as their essential feature comparison of outcomes between a 'treatment' group who have participated in, or been directly affected by, a government program or policy, and a 'control' group of non-participants. Possible applications of the methodology extend to virtually every area of government activity. International studies have, for example, examined how participation in labour market programs affects subsequent employment and earnings outcomes; effects of anti-poverty programs; effects of compulsory schooling attendance legislation; whether minimum wages affect employment; and effects of mandated maternity benefits on female employment and earnings.

The importance of the research area of program evaluation derives from its value as an empirical methodology for understanding effects of government policy; and the increasing demand for program evaluation that is likely to come from governments in Australia in the next 5 to 10 years.

Good policy-making must be informed by knowledge about the predicted effects of proposed policies. An understanding of predicted effects can come from empirical analysis of the effects of previous or existing policies that are similar to the proposed policy, or from theoretical or simulation modelling of the effects of the proposed policy. It is now well-accepted that the most valuable methodology for analysis of effects of previous or existing policies is program evaluation (experimental and quasi-experimental methods). It has several main advantages. First, it is highly flexible in the range of types of programs or policies to which it can be applied. Second, compared to alternative approaches such as regression modelling, much less restrictive assumptions on the nature of the program and its effects are required to be able to derive valid estimates of the program effect. (For example, the program evaluation approach can more easily accommodate heterogeneity in the effect of a program across the population who participate in the program.)

That good policy-making should be 'evidence-based' does seem to have been increasingly recognised by governments in recent times. Hence, a technique such as program evaluation - that provides direct evidence on the effect of a program on specified outcomes that relate to its objectives - can only become more important and influential. The international experience has been that governments have over the past 10-20 years significantly expanded the resources they have committed to program evaluation, and that the findings from such evaluations have been critical in policy-making. In Australia, there has been some move in the same direction - through for example the research program of the Commonwealth Department of Family and Community Services, and work undertaken by state Workers' Compensation agencies. Based on international experience, it seems highly likely that there will be a significant increase in demand for research on program evaluation in future years.

Projects commenced in 2010

Projects commenced in 2008

Projects commenced in 2006

Projects commenced in 2005

Projects commenced in 2004

Projects commenced in 2001