Investigating the utility of indexes in comparative drug policy analysis

Abstract

Drug policy is multi-dimensional and complex. It is also undergoing dramatic shifts; away from prohibition toward more diverse policy approaches. It is therefore a crucial time to develop new measures and ways to evaluate the impact of these policy changes. One way of doing this is through comparative policy analysis (CPA). The method specifically explored here is the development of composite indicators (indexes) with a focus on “laws on the books”. Indexes are created by combining multiple dimensions of a policy into a single score. This thesis explores the methodological challenges inherent in index design including; deciding what is included in the index, normalising all values to create a common metric, weighting, and aggregating the indicators. Two indexes were developed in order to explore index development as a way to measure drug laws and to then assess how they may be used in CPA. An index of Australian Cannabis Laws and Regulations (the ACLRI) measured the level of severity of cannabis control laws and regulations across eight states and territories. Higher index scores reflected more severe cannabis control laws. The relationship between the ACLRI and the prevalence of cannabis use and harms were assessed. The prevalence of cannabis use was not related to index score, but an increase in index score was potentially related to increasing harmful consumption. An Opioid Overdose Prevention Index (OOPI) quantified the introduction and effectiveness of opioid overdose prevention laws concerned with naloxone access and Good Samaritan provisions across 51 US states between 2000 and 2016. One sub-index was found to be valid. Regression analysis testing the relationship between the OOPI score and the US opioid overdose mortality rates suggested that increased law introduction was not associated with reductions in opioid overdose rates. This is the first application of index methods to compare drug laws. Indexes were found to be useful for capturing complexity and revealed diversity in state legislative approaches and related outcomes when applied to CPA such as the finding that differences in cannabis law severity may have an impact on harmful consumption. Nevertheless, the key finding that laws appeared to have little effect on drug policy outcomes reinforces that while laws may be necessary, they may not be sufficient to achieve policy goals. Future drug policy indexes may thus benefit from incorporating measures of implementation.