1500.0 - A guide for using statistics for evidence based policy, 2010  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 20/10/2010  First Issue
   Page tools: Print Print Page Print all pages in this productPrint All  
Contents >> Evaluate outcomes of policy decisions


Once a policy has been implemented it is necessary to monitor and evaluate the effectiveness of the policy to determine whether it has been successful in achieving the intended outcomes. It is also important to evaluate whether services (outputs) are effectively reaching those people for whom they are intended. Statistics play a crucial role in this process. As Othman (2005) states, ‘Good statistics, therefore, represent a key role in good policy making. The impact of policy can be measured with good statistics. If policy cannot be measured it is not good policy.’

Performance indicators can be used to evaluate policy, and these should be supported by timely data of good quality. Measuring Wellbeing, 2001 (cat. no. 4160.0) provides a list of elements that effective indicators should include:

    Relevant/ reflective of issue: indicators need to reflect the social, economic or environmental issues, or alternatively, the policy decisions of government.

    Available as a time series: data about a single point in time can allow for comparison between population groups or geographical areas. However, when the indicator can be repeated at a later time, change in the phenomenon can be assessed.

    Meaningful and sensitive to change: a successful indicator needs to closely reflect the phenomenon it is intended to measure and be realistic. It needs to relate to associated measures in a logical way and should ideally respond to changes in the real world. For example, changes in average population height will not directly or quickly reflect changing levels of nutrition.

    Summary in nature: a large mass of information can be represented by a few indicator time series that bring out the main features of the issue, e.g. female to male earnings ratios. In some circumstances however, the summary nature of indicators can be misleading, e.g. a low perinatal mortality indicator for the total population can mask the higher perinatal mortality rates of Australia’s Indigenous population.

    Able to be disaggregated: indicators must not only reveal national averages but be capable of finer division, e.g. many social indicators vary sharply by age and sex.

    Intelligible and easily interpreted: indicators need to be readily understood and not overly complex. It should be obvious exactly what an indicator is showing, and how it can be applied in practice. Life expectancy tables are often based on complex statistical and actuarial techniques, however they can be readily understood by the public.

    Able to be related to other indicators: a variety of indicators can support a central measure or show relationships and interdependencies.

Previous PageNext Page