While randomized controlled trials (RCTs) are the 'gold standard' for impact evaluation, they face numerous practical barriers to implementation. In some circumstances, a randomized-encouragement design (RED) is a viable alternative, but applications are surprisingly rare. We discuss the...

While randomized controlled trials (RCTs) are the "gold standard" for impact evaluation, they face numerous practical barriers to implementation. In some circumstances, a randomized-encouragement design (RED) is a viable alternative, but applications are surprisingly rare. We discuss the...

This paper considers the degree to which events that intensify partying increase sexual assault. Estimates are based on panel data from campus and local law-enforcement agencies and an identification strategy that exploits plausibly random variation in the timing of Division 1 football games....

We seek to quantify the role of education as a mechanism through which family background affects earnings. To this end, we propose a generalisation of statistical 'mediation analysis'. In our approach, the treatment and mediator can be multidimensional. This allows us to directly and flexibly...

This paper aims to shed new light on explanations for the sexual division of labour, within a broader examination of within-household specialisation. We propose a set of indices which we believe are the first direct within-couple measures of specialisation. We use these to present a rich...

Climate change has stimulated growing interest in the influence of temperature on cognition, mood and decision making. This paper is the first investigation of the impact of temperature on the outcomes of criminal court cases. It is motivated by Heyes and Saberian (2019, AEJ: Applied Economics),...

This paper studies how having your home damaged or destroyed by a natural disaster impacts on economic and financial outcomes. Our context is Australia, where disasters are frequent. Estimates of regression models with individual, area and time fixed-effects, applied to 10 waves of data...

We propose a new model-selection algorithm for Regression Discontinuity Design, Regression Kink Design, and related IV estimators. Candidate models are assessed within a 'placebo zone' of the running variable, where the true effects are known to be zero. The approach yields an optimal...

We present the first estimates of intergenerational wealth correlation for Australia, using HILDA. The rank correlation varies greatly by child age when wealth is observed, from 0.1 before age 30, to 0.5 after age 40. Most children in our estimation sample are young. For these children overall,...