Policy on trial

Source: Policy, Spring 2008
"Evidence-based policy is all the rage in Canberra, yet policymakers are not testing their ideas to see if they will work before rolling them out, unproven, at taxpayers’ expense. This author argues that randomised trials are the epitome of rational inquiry: they can establish whether a causal effect has occurred and thus prove whether government action is effective or not. Government policies that would be good candidates for randomised trials include the responses to the 2020 Summit and testing whether the provision of high-speed internet access to schools will benefit students and increase their grades. Yet, randomised trials are politically unattractive. Stating a hypothesis can be problematic for policymakers, forcing them to move from vague statements of intent to a specific measurable outcome they wish to achieve. On top of this, making a real difference is hard and randomised trials can show that government intervention makes no real difference."