Research
Research
with Alan Duncan, Astghik Mavisakalyan and Lili Loan Vu
An under-explored aspect of body weight is the role that location plays in its determination. To address this evidence gap, this study uses a dynamic event study framework to examine how weight changes for people who change their location and, by extension, how much of the overall variation in weight can be explained by place-based versus individual factors. We then undertake a decomposition analysis on movers and non-movers to determine the role of place in differences in weight between groups of areas. We also model the effects of place on common behavioural correlates of weight. Our results, based on a rich longitudinal survey from Australia, indicate a significant robust place effect on weight. Our event study shows that place is responsible for 15.5% of the variation of weight. We find similarly sized place effects when comparing different groups of areas in our decomposition exercise. Finally, we show that place exerts a strong influence, of around 50% in some cases, on food consumption.
Paper presented at ESPE 2024 conference in Rotterdam, EuHEA 2024 conference in Vienna, ICEE 2024 online conference hosted by PSU Altoona, the Global GLO-JOPE online conference 2024, and a milestone seminar at Curtin University in Perth.
with Alan Duncan, Astghik Mavisakalyan and Lili Loan Vu
Air pollution affects 99% of the global population at levels exceeding WHO guidelines, yet its impact on governance and civic engagement remains largely unexplored. We combine three waves of the Life in Transition Survey across 29 countries (2006-2016) with high-resolution pollution and meteorological data to examine how air pollution affects corruption perceptions and political participation. Using temperature inversions as instrumental variables for PM2.5 exposure, we find that each additional μg/m³ increases perceived petty corruption by 1.8% overall and 6.4% in EU transition countries, while increasing the odds of perceiving government officials as corrupt by 6-8%. Air pollution also reshapes political participation. Each additional μg/m³ reduces voting by 8-10% while increasing petition signing by 10% in Central and Eastern European countries. This suggests that pollution drives substitution from institutional to non-institutional engagement. Effects concentrate among older, less-educated, and male populations.
Paper presented at the WA AARES 2025 Conference in Perth and Global GLO-JOPE Online Conference 2025
Conflict can shape multiple behaviours and beliefs amongst those exposed to it - with one of the most important factors being their religious beliefs and behaviours. Currently there is a lack of research into the long-term impacts of conflict on religious behaviour, which our research seeks to address in the context of post-war Japan. Combining urban bombing data from the United States Strategic Bombing Survey with pre-war Japanese census data and the Japan General Social Survey we exploit variations in bombing intensity and differing exposure by birth cohorts to conduct a difference-in-difference analysis to examine the degree to which exposure to strategic bombing in the Second World War impacts later life religious practice amongst Japanese citizens. To account for issues of endogeneity in the targeting of bombing we use a novel instrument; the deviation between planned and actual square kilometres destroyed, to conduct a two-stage least squares regression. We find that increased exposure, concentrated amongst exposure in childhood and adolescence, to strategic bombing decreases later life participation in religious practice. For each additional standard deviation of damage individuals are 1.86% less likely to practice religion in later life. These results are robust to multiple definitions of bombing intensity, age exposure bands and sample specification. We also find evidence of negative effects on faith amongst those exposed in adolescence with a 1.9% decrease in faith for each additional standard deviation in damage. These findings suggest that conflict can impact both religious practice and belief in the long-term.
Paper to be presented at NEUDC 2026 at Cornell University