A Deakin workshop is set to host applications of an important research methodology that’s also been recognised in this year’s Nobel Prize.
Tackling some of the big questions in social and economic sciences, “natural experiment” is a method that’s applied to understand the effect that sudden developments – such as changes in policy, legislation, borders, or wars, conflicts, terrorist attacks, floods, bushfires, and pandemics – have across the domains of health, education, development, labour and political economics.
Professor Mehmet Ulubasoglu, who heads up DBS’s Department of Economics, says that natural experiments are similar to randomised control trials (RCT) but instead, use situations where interventions from nature, rather than from the researcher, form the treatment and control groups which helps understand the impact of the intervention.
‘For example, we may want to understand the effect that malnutrition has on the long-term health or education outcomes of children. Of course it’s highly unethical to apply a lab experiment or RCT on children and keep them malnourished, but we can look at world-wide famines in history and investigate the effects – as a “natural” experiment of malnutrition – on the affected cohort of children vis-à-vis unaffected cohorts. Natural experiments are called quasi-experimental methods,’ he explains.
Prof. Ulubasoglu suggests that over the past three decades, natural experiment has grown into a mainstream research approach.
‘The method enables researchers to pin down the causal effects in economic relationships, hence separate correlation from causation. Most recently, its prominence has been recognised by the Nobel Prize in Economic Sciences 2021 which was awarded to David Card (UC Berkeley), Joshua Angrist (MIT) and Guido Imbens (Stanford) who, over the past 25 years have widely disseminated its use in economics,’ he says.
Using natural experiments, Prof. Card has analysed the labour market effects of minimum wages, immigration and education and since the 1990s, his studies have challenged conventional wisdom, leading to new analyses and additional insights that help policy-makers prepare and plan for the future.
‘Angrist and Imbens’ contributions transformed the understanding of econometric models in that we teach quantitative methods differently now. We place more emphasis on causal effect in undergraduate and doctoral programs,’ says Prof. Ulubasoglu.
Widely-recognised for his research into the economic impact of natural and man-made disasters, Prof. Ulubasoglu has applied natural experiments to a broad scope of significant projects.
As well as a recent article in Health Economics on the long-terms effects of the Great Vietnam Famine, he has a newly published article in the Journal of Economic Behavior and Organisation (with David Johnston (Monash), Yasin K Onder (Ghent), Md Habibur Rahman (Curtin)), which evaluates wellbeing losses due to the 2009 Black Saturday Bushfires using natural experiment methodology.
‘My four BNHCRC (Bushfire and Natural Hazards CRC) research reports on the income effects of Black Saturday Bushfires 2009, Queensland Floods 2010/11, Cyclone Oswald 2013 and Toodyay Bushfires 2009 all use the natural experiment methodology. Notably, this work has been discussed in the CSIRO 2020 Report on Climate and Disaster Resilience submitted to Prime Minister Scott Morrison,’ he says.
Hosted by DBS and the Alfred Deakin Institute, the 8th Annual Workshop on Natural Experiments in History: Development, Health and Labour will be presented on 8 December via Zoom, hosting nine speakers including keynote speaker Professor Ahmed Mushfiq Mobarak (Yale University) who is also a professorial fellow with the Department of Economics at Deakin University.
Prof. Ulubasoglu says he and former Deakin colleague Dr Cahit Guven long ago realised the importance of natural experiment methodology and established the workshop to bring together Australia-wide researchers, PhD students and international keynote speakers.
‘Australia is home to many strong empirical researchers and every year several PhD theses are completed using natural experiments. This year’s organising committee includes junior colleagues Drs Ha Vu and Lan Anh Tong. Natural experiment methodology is a neat way of measuring the causal economic impacts of natural or manmade shocks or abrupt changes that occurred in history on an array of short - and long-term economic outcomes. This helps formulate new policies or fine-tune existing policy prescriptions.’
Information on previous events (including previous workshops and programs) can be found here.