Publications

Bernice Hua Ma, Samia Badji, Dennis Petrie, and Gang Chen

2025-19

This study examines the impact of enhanced social care provided through the Australian National Disability Insurance Scheme (NDIS) on subsided healthcare utilisation for people with disability. Using linked administrative datasets from 2011 to 2020, we employed a Difference in Difference model and the staggered rollout of the NDIS to assess its effects on healthcare services, focusing on visits to general practitioners (GP), mental healthcare providers, allied health professionals, specialists, and mental health prescriptions. The results show that the NDIS reduced subsidised mental health services and allied health services in the six quarters after enrolment. However, it did not significantly affect visits to GP, specialists, or mental health prescriptions. These effects were most pronounced among individuals aged 0-24 years, males, and those living in major cities. The findings suggest that services available from NDIS may substitute for subsided healthcare services by providing non-clinical care through social care channels. Further research is needed to investigate the long-term effects and health outcomes of the NDIS.

Read the paper

Adam Irving, Dennis Petrie, Anthony Harris, Laura Fanning, Erica M. Wood, Elizabeth Moore, Cameron Wellard, Neil Waters, Bradley Augustson, Gordon Cook, Francesca Gay, Georgia McCaughan, Peter Mollee, Andrew Spencer, Zoe K. McQuilten

2025-18

Background

Health technology assessments traditionally rely on cohort modelling using clinical trial data, leaving uncertainties about real-world cost-effectiveness. This post-market economic evaluation used individual-level modelling with a discrete-event simulation (DES) framework and registry data to estimate the real-world cost-effectiveness of bortezomib, lenalidomide and dexamethasone (VRd) in Australia which was listed for newly diagnosed multiple myeloma in 2019.

Methods

We conducted an economic evaluation of VRd versus No VRd using the EpiMAP Myeloma model, a DES model powered by risk equations from the Australia & New Zealand Myeloma and Related Diseases Registry. This approach captured individual patient heterogeneity and complex treatment pathways through up to nine lines of therapy. We assessed differences in quality-adjusted life-years (QALYs) and costs over a lifetime horizon, with bootstrapping to quantify uncertainty.

Results

VRd was associated with positive incremental QALYs (0.16; 95% CI: 0.10, 0.21) and incremental cost (A$10K; 95% CI: A$8K, A$11K). Improved response to first-line therapy with VRd was predicted to marginally increase receipt of autologous stem cell transplantation by 1.1% (95% CI: 0.6, 1.7%), significantly increase receipt of maintenance therapy by 13.8% (95% CI: 10.4%, 17.3%) and marginally offset further lines of therapy. VRd was the most cost-effective option in 95% of the bootstrap iterations at a willingness-to-pay threshold of $60K/QALY.

Conclusion

The 2019 decision to list VRd for newly diagnosed multiple myeloma has resulted in a somewhat cost-effective allocation of healthcare resources when judged against the traditional A$50K/QALY willingness-to-pay threshold. This analysis demonstrates how using individual-level modelling with registry data to perform economic evaluation can capture the interplay between patient characteristics, treatment decisions, and outcomes. Our findings provide nuanced insights into the real-world cost-effectiveness of VRd, highlighting how post-market evaluations can inform refinement of funding decisions for complex therapeutic interventions.

Read the paper

Adrianna Bella, Nicole Black, Teguh Dartanto, Danusha Jayawardana, and Dennis Petrie

2025-17

Child marriage remains a significant global issue, violating human rights and limiting development outcomes, particularly for girls. This study examines the impact of Indonesia’s first minimum marriage age (MMA) law, which in 1975 set the minimum age of marriage for girls at 16 years. The analysis relies on a regression discontinuity design to estimate the effects of the policy on child marriage and girls’ education, with specific adjustments to address non-random heaping in reported years of birth. Using data from the 2018–2021 Indonesia National Socio-economic Survey (SUSE-NAS), we find that the MMA law reduced marriages under the age of 16 by 18% and increased the age at first marriage by about five months. It also had a broader effect by delaying marriages beyond the legal threshold. The effects were stronger in regions with entrenched child-marriage norms and in urban settings. We also find evidence that the MMA policy had positive effects on educational attainment, particularly in obtaining a tertiary degree.

Read the paper

Jessica Arnup, Nicole Black, and David Johnston

2025-16

Read the paper

Ashani Abayasekara, Sonja de New, and David Johnston

2025-15

Read the paper

Anthony Scott, Susan Méndez, Jongsay Yong

2025-14

This paper examines the impact of a fall in demand for voluntary private health insurance on physician behaviour. We find that dual practice physicians who earned more revenue from working in private hospitals before the fall in demand for PHI were more likely to experience a fall in the volume of private hospital care and a less complex mix of in-hospital services provided. Risk-averse doctors drove this reduction in complexity. There was weak evidence suggesting that doctors compensated for the fall in volume by increasing their working hours in public hospitals and reducing fees for in-hospital services. We found no evidence of higher volumes of care in out-of-hospital visits to compensate for the lower in-hospital volumes, but some weak evidence of claiming for more complex and costly out-of-hospital services, which we interpret as upcoding. Compensatory behaviours were small and primarily observed among risk-averse male doctors with low conscientiousness, from non-surgical specialties, and located in areas with lower doctor density.

Read the paper

Nicole Black, Danusha Jayawardana, David W. Johnston and Trong-Anh Trinh

2025-13

Rising out-of-pocket costs for psychotherapy in Australia have heightened concerns about financial barriers to mental healthcare, particularly for lower-income households, who disproportionately experience psychological distress. Using nation-wide linked administrative records of income and healthcare use, we estimate the magnitude of income-related inequity in psychotherapy use among 5.4 million individuals diagnosed with a mental health condition, and examine how such inequity has evolved over the decade from 2014 to 2023. Our findings show that income-related inequity is substantial, consistently higher among children than among adults, and has nearly doubled over the decade. By 2023, only 32% of low-income children and 40% of low-income adults accessed psychotherapy within three months of receiving a mental health treatment plan, compared with 55% among both high-income children and adults. We rule out changes in complexity of mental health disorders and the introduction of telehealth services as key drivers. We find no discernible difference by gender or age subgroups. Examination of antidepressant use reveals a growing gap in the opposite direction, with lower-income individuals increasingly reliant on medication without psychotherapy, relative to higher-income individuals. This suggests a shift towards lower-cost treatment pathways among disadvantaged groups. Our findings highlight the need for policies to address the increasing costs and other barriers to accessing psychotherapy, especially for lower-income households.

Read the paper

Nicole Black, Danusha Jayawardana, and Gawain Heckley

2025-12

Recent research shows that birth order affects human capital outcomes, yet there is limited empirical evidence on the underlying mechanisms. This study examines the effect of birth order on children’s time use across activities that are important for human capital development. Using detailed time-use diaries of Australian children aged 2-15, we find that within families with two or three children, later-born children spend less time on enrichment activities and more on digital media, compared to first-born children. We obtain the same findings when we repeat the analysis using detailed time-use diaries of US children. Further investigation reveals that part of the birth order effect is driven by parents spending less time with later-born children compared to first-borns. However, later-borns also independently devote less of their own time to enrichment activities, suggesting that personal time use may be an important mechanism behind the well-documented impact of birth order on human capital development. We find evidence that later-born children experience more lenient parenting, which may help explain this pattern of own time use.

Read the paper

Karinna Saxby, Thomas Buchmueller, Sonja C. de New, Dennis Petrie

2025-11

Poor mental health is a major global health issue, with many countries documenting high levels of unmet need and regional disparities in mental healthcare utilization. To determine how best to address these disparities, it is important to understand what drives regional variation. Using Census-linked microdata from Australia, we exploit cross-region migration to identify the extent to which patient and place factors drive regional variation in utilization of mental healthcare services and mental health prescriptions (antidepressants, anxiolytics, antipsychotics). We find that place factors account for approximately 72% and 19% of the regional variation in utilization of mental healthcare services and mental health prescriptions, respectively, with the rest reflecting patient-related demand. We also find suggestive evidence that larger place effects predict fewer mental health related ED presentations, self-harm hospitalizations, and suicides. Altogether, our findings suggest there is inadequate and inequitable supply in regions with low utilization, rather than inefficiently high utilization in high utilization regions.

Read the paper

Daniel Avdic,  Bo Lagerqvist, Nils Gutacker, Giovanni van Empel, Johan Vikström‖

2025-10

We study how multi-technology hospitals respond to market entry of single-technology competitors using a rescindment of regulations for heart attack treatments that prompted a rapid expansion of catheterization laboratories (cath labs) in Sweden. We isolate supply-side effects by exploiting that patients cannot choose their hospital and compare outcomes of cardiac patients residing in areas affected and unaffected by provider market entry, respectively. We show that patients with indications for cardiac surgery were more likely to receive catheter-based treatment after a cath lab opened in their hospital, and document increases in adverse health outcomes for inframarginal patients. Incumbent hospitals responded to this demand reallocation by augmenting their own demand for surgery, but to a lesser extent and without patient health consequences.

Read the paper

Jing Jing Li, Anthony Harris

2025-09

Public funding decisions for pharmaceuticals are the outcomes of a dynamic bilateral bargaining process between the funding agency and a company and can involve considerable delay. Using an empirical duration model of negotiation in Australia from 2005 to 2018, we test if agreement patterns on national public subsidy of pharmaceuticals are consistent with the predictions of dynamic bargaining theory. It took a median of 16 months for the Australian government and companies to reach an agreement, averaging 1.51 rounds of negotiations, with 71% of the rounds failing to reach an agreement. Overall, the results of a process of one-sided offers from companies are consistent with theories of bargaining with incomplete information and delay strategies, where evidence of quality develops over negotiation rounds. Lower value and more risk for the payer delayed agreements and increased the probability of no agreement, while public awareness and interest in a drug reduced the agency’s bargaining power and increased agreement rates. Enhanced knowledge about the drug’s attributes benefits the government and its constituencies, but pharmaceutical companies have a strong incentive to invest in political alliances to raise awareness of potential benefits to patients and hasten public funding of a drug.

Read the paper

Hai-Anh H. Dang, Stephane Hallegatte, Minh Cong Nguyen, Trong-Anh Trinh

2025-08

Despite a vast body of literature documenting the harmful effects of climate change on various socio-economic outcomes, little cross-country analysis exists on the global impacts of higher temperatures on poverty and inequality. Analyzing a new global panel dataset of subnational poverty in 137 countries covering the past decade, we find that a one-degree Celsius increase in temperature leads to a 17.1% increase in poverty, employing the US$2.15 daily poverty threshold, and a 1.1% increase in the Gini inequality index. We also find negative effects of colder temperature on poverty and inequality. Yet, while poorer countries—particularly those in Sub-Saharan Africa—are more affected by climate change, household adaptation could have mitigated some adverse effects in the long run. The findings provide relevant and timely inputs for the global fight against climate change as well as the current policy debate on cost-sharing between richer and poorer countries.

Read the paper

Preety Srivastava, Trong-Anh Trinh, Xiaohui Zhang

2025-07

This paper contributes to the growing body of research examining the impact of temperature on educational outcomes. Utilising national-level administrative data on nearly one million Australian students, it investigates whether temperature fluctuations, and prolonged heatwaves influence test performance. The analysis reveals that both heat and cold affect student test scores, with some evidence of the effects intensifying during heatwaves. Australia’s vast geographical diversity and climate variability provides a unique opportunity to explore spatial heterogeneity in these effects. Findings suggest that in regions with hot weather conditions, the most thermally comfortable temperature is likely to be higher, whereas students in the coldest parts of the country appear to be less sensitive to cold weather conditions, consistent with the adaptation hypothesis. In contrast, in regions with moderate and temperate weather conditions, student scores are affected by both hot and cold weather.

Read the paper

Sefa Awaworyi Churchill, Simon Chang, Russell Smyth, Trong-Anh Trinh

2025-06

This paper extends prior theory linking present-day sex ratios to present-day propensity for entrepreneurship among men backward in time to explore the long-run gender origins of entrepreneurship. We argue that present-day propensity for entrepreneurship among men will be higher in neighbourhoods which had historically high sex ratios. We propose that high sex ratios generate attitudes and behaviours that imprint into cultural norms about gender roles and that vertical transmission within families create hysteresis in the evolution of these gender norms. To empirically test the theory, we employ the transport of convicts to the British colonies of New South Wales and Van Diemen’s Land in the eighteenth and nineteenth centuries as a natural experiment to examine the long-run effect of gender norms on entrepreneurship in present-day Australia. We use a representative longitudinal dataset for the Australian population that provides information on the neighbourhood in which the participant lives, which we merge with data on the sex ratio in historical counties from the mid-nineteenth century. We find that men who live in neighbourhoods which had high historical sex ratios have a higher propensity for entrepreneurship. We present evidence consistent with the vertical transmission of gender norms within families being the likely mechanism. Arguments for policies to promote female entrepreneurship are typically couched in terms of gender norms representing a barrier to more women starting their own business. We present evidence consistent with gender norms contributing to gender differences in rates of entrepreneurship by being a spur for higher male entrepreneurship rather than a barrier to female entrepreneurship.

Read the paper

Sander de Vries, Nadine Ketel, Maarten Lindeboom

2025-05

There is a clear consensus that childhood experiences shape adult success, yet there is limited understanding of their impact on future generations. We proxy parental investments during childhood with birth order and study whether disadvantages due to lower investments are transmitted to future generations. Birth order effects on the first generation are large, apply to 80% of the population, and can be identified with relatively mild assumptions. Using cousin comparisons in Dutch administrative data, we find that around 20 percent of the income disadvantages are transmitted. Additionally, we find sizeable decreases in children’s education and increases in boys’ criminal behaviour.

Read the paper

Gawain Heckley, Dennis Petrie

2024-18

This paper presents a flexible method, Parameter Estimation by Raw Moments (PERM), to evaluate a policy’s impact on parameters of the distribution of outcomes. Such parameters include the variance (E[Y2]-E[Y]2), skewness and covariance. While many studies estimate the mean (first moment), PERM extends this to estimate higher order moments, enabling calculation of distribution parameter treatment effects. Two implementations are discussed: regression with controls and DiD with staggered rollout. Applying PERM DiD to a Swedish school reform shows it reduced education inequality but increased earnings variance resulting in a lower covariance between education and earnings.

Read the paper

Nicole Black, Anthony Harris, David W. Johnston, and Trong-Anh Trinh

2024-17

To improve access to affordable mental healthcare, Australia introduced a national, publicly-funded program in 2006 to subsidise psychological services, which were previously privately funded. Theory suggests that by increasing demand, subsidies should incentivise providers to expand supply, either by entering the workforce or by increasing hours. However, expansion depends on supply constraints, particularly in non-metropolitan areas with pronounced workforce shortages. This study examines the program’s effects on the size, work hours, demographic composition, earnings, and location of practising psychologists. Using a difference-in-differences approach with census data and Australian Tax Office records, we find that the program: (1) increased the supply of practising psychologists by 47% within five years, with minimal effect on average hours worked per psychologist; (2) raised psychologists’ annual personal incomes by 17%; and (3) increased the likelihood of psychologists working in non-metropolitan areas by 8%, supporting broader access to mental healthcare.

Read the paper

Muhammad Fikru Rizal, David W. Johnston, Nicole Black, Rohan Sweeney

2024-16

In many low and middle-income countries (LMICs), a high proportion of people with mental health needs do not receive treatment, contributing to a significant “treatment gap”. Despite this, there is limited robust evidence on the socioeconomic factors that shape mental healthcare use in these settings. Using data from over 400,000 adults in Indonesia, this study examines how wealth, education, and health insurance coverage influence the likelihood of accessing mental healthcare among those with probable depression. Indonesia is an important context for this analysis because undertreatment and stigma are particularly severe. We find that only 9.3% of those identified as having probable depression receive treatment. Wealth and health insurance are positively associated with the probability of mental healthcare utilisation, while education is not. The wealth gradient diminishes at the highest income levels, and we show this is possibly due to increased stigma. These findings underscore the importance of reducing financial barriers, such as through public health insurance expansion and reducing stigma to address the mental health treatment gap in LMICs.

Read the paper

Rosanne Freak-Poli, Stephen P. Jenkins, Michael A. Shields, and Trong-Anh Trinh

2024-15

Despite a substantial literature on the links between social relationships and mortality, the size of the relative risks from loneliness, social isolation, and living alone, remain controversial. Further research is therefore important given demographic changes meaning that more people are living alone, for longer, and with chronic health conditions. Using 19 waves of high-quality Australian longitudinal data we provide new evidence using multiple measures of social relationships, model specifications, and adjustments for confounding. We focus on chronic measures of (poor) social relationships and provide separate estimates by gender. We find that both functional and structural aspects of social relationships are independently strongly associated with all-cause mortality. We estimate a hazard ratio for loneliness of 1.41, which is greater for males (1.55) than females (1.24). These hazard ratios are larger than found for social isolation (1.19). We also find a strong relationship between being an active member of a club and reduced mortality risk, but no evidence that living alone is an independent risk factor. We provide useful comparisons with the mortality risks associated with smoking and household income. Overall, our findings suggest that interventions should focus on reducing both loneliness and social isolation, as well as encouraging active social participation.

Read the paper

Nicole Black, David W. Johnston, Martin Knapp, Michael A. Shields, and Gloria H.Y. Wong

2024-14

For people experiencing mental health problems, timely access to high-quality healthcare is imperative for improving outcomes. However, limited availability of services, high out-of-pocket costs, insufficient health literacy and stigmatising attitudes may mean people do not receive the necessary treatment. We analyse Australian longitudinal data to document the extent and predictors of horizontal inequity in mental healthcare use among people with a newly developed mild or moderate mental disorder. Importantly, we compare people with similar health, residing in the same area, thus controlling for differences in healthcare needs and availability of services. Results suggest that mental healthcare use is not significantly associated with household income or financial hardship. In contrast, we find significant inequities by educational attainment, with university graduates around 50% more likely to receive mental healthcare than high-school dropouts. These findings are robust across subsamples and alternative modelling approaches, including panel data models with individual fixed- effects. Additional explorations of the education gradient suggest a potential pathway through mental health-specific knowledge and attitudes.

Read the paper

Miao Guo, Yang Li, Minghao Wu and Terence C. Cheng

2024-13

This study analyzes the effects of China’s long-term care insurance (LTCI) benefit design on household consumption and intergenerational support. The program provides two benefit options: in-kind benefits (or services) and cash allowances. We introduce a conceptual framework to analyze economic decision making under the two types of LTCI benefits. Using an empirical framework that exploits variations in LTCI benefit designs across China’s pilot cities, we find that both types of LTCI benefits increase household consumption and reduce medical expenditure. Specifically, ‘mixed’ benefits households – those with a choice between in-kind and cash benefits – significantly increase spending on food and housing, while households receiving services spend more on housing, transport, and clothing. Additionally, in-kind benefit recipients report receiving lower informal care from their children, implying a substitution with formal care. Households with mixed benefits experience a decline in financial support from children, suggesting a crowding-out of intergenerational transfers. Finally, we estimate income and substitution effects that are implicit in recipients’ behavior to analyze welfare implications under China’s LTCI.

Read the paper

Nicole Black, David W. Johnston, Michael A. Shields, and Trong-Anh Trinh

2024-12

We study the extent of horizontal inequity in children’s mental healthcare use in Australia, where universal insurance aims to provide equitable access to needed treatment, regardless of ability to pay. We use linked longitudinal survey data and administrative records that measure the need for mental healthcare – via screening questionnaires and general practitioner (GP) diagnosis – and use of mental health professionals and medication. Using between- and within-child approaches, we find that conditional on need, children from lower income families are significantly less likely to receive services from clinical psychologists than children from higher income households. However, we see little evidence of income inequities in receiving mental health services from GPs or general psychologists. We show that differences in out-of-pocket fees are a likely explanation. The findings highlight that specific support to low-income families is needed to reduce inequities in accessing a complete range of mental health services.

Read the paper

Nicole Black, Johannes Kunz

2024-11

Language proficiency is a crucial skill for immigrants that influences their social integration and their children’s development. This study examines the intergenerational effects of limited English proficiency (LEP) on children’s health and health care utilisation. We mitigate potential selection issues arising from insurance coverage by examining Australian-born children who are all covered under a universal public health insurance scheme. We use Australian population Census and longitudinal survey data linked to administrative health care records, and variation in parent’s language acquisition, based on their age at arrival into Australia. We find that parental LEP has a strong and positive effect on children’s use of general practitioners, but no effect on their use of other healthcare services, or on their physical or mental health. We explore several possible supply- and demand-side explanations.

Read the paper

Michelle Escobar Carías, Nicole Black, David Johnston, Rohan Sweeney, Fiona S. Barker, Rosnaena, Syaidah Syamsul, Taniela Waka

2024-10

We document gender differences in children’s time investments in education, labour, and leisure in an understudied population of children living in urban informal settlements. Using within-settlement and within-sibling comparisons, we find that boys spend significantly less time than girls on schooling and homework and more time on leisure activities. We also find that caregivers invest less time in helping their sons with reading and homework than their daughters. One possible explanation is that girls spend more time on domestic work. As a result, as the share of girls in the household increases, primary caregivers spend less time on domestic work and more time on other activities such as teaching children. We find that the gender gaps in time use are more pronounced among children whose parents have lower schooling and more financial constraints.

Read the paper

Daniel Avdic, Johannes S. Kunz, Susan J. Méndez, Maria Wiśniewska

2024-09

We study how the diffusion of telemedicine technology impacted the quality and rates of antibiotic prescriptions using Australian survey data from primary care physicians linked to administrative records on their service provision. We classify physicians based on their relative use of telemedicine consultations in response to the introduction of government-subsidised telemedicine during the COVID-19 pandemic and relate their rates of antibiotic prescriptions to indicators of pre- scribing quality before and after lockdown periods in a difference-in-differences design. Our results suggest that more frequent users of telemedicine prescribe relatively fewer antibiotics while keeping prescribing quality largely unchanged. We interpret these findings as evidence that telemedicine can enhance efficiency of service provision in primary care settings.

Read the paper

Wang-Sheng Lee, Umair Khalil, David Johnston

2024-08

This paper estimates the impacts of religiosity on criminal activity using a city-wide shock to religious sentiment from a 2015 Papal visit. Using daily data on all reported offences between 2010 and 2015 in Philadelphia at the census tract level and a difference-in-differences approach, we demonstrate significant reductions in less serious crimes in the week of the visit and for several weeks following. Reductions are particularly pronounced for drug offences and in historically Christian areas. Notably, similar crime effects are not found for President Obama’s 2015 visit, suggesting changes in police deployment do not drive results.

Read the paper

Johannes Kunz, Carol Propper, Trong-Anh Trinh

2024-07

Digital access may bring important health gains, particularly where physical infrastructure is limited. We examine the impact of internet access in Indonesia on health outcomes using the COVID-19 pandemic as a health shock. We utilize sub-national data on mobile broadband, COVID-19 spread, and an instrumental variable approach using lightning strikes as an exogenous shock to connectivity. Access to 3G internet significantly reduced the transmission of COVID-19. Areas with internet access had approximately 45% fewer cases. Regions with higher literacy and capacity for telework benefited significantly more. These findings offer novel insights into how digital infrastructure affects public health outcomes.

Read the paper

Stephen Hoskins, David W. Johnston, Johannes S. Kunz, Michael A. Shields, Kevin E. Staub

2024-06

Despite being widely used in health economics, dynamic models of health and healthcare typically assume that the persistence in these outcomes is the same for every individual. Understanding the extent and drivers of heterogeneity in persistence is essential for the design and evaluation of health interventions because persistence determines the dynamics and overall long-run effects of such interventions. This paper explores individual-level heterogeneity in the persistence of health outcomes. Using simple regression methods that do not place any restriction on the distribution of the heterogeneity in persistence, the paper documents substantial heterogeneity in health, medical expenditures, and healthcare use. We show that neglecting this heterogeneity leads to estimates that overstate the average persistence and can bias the coefficients of covariates. We find that between 75% and 87% of individuals display persistence and that this persistence is related to the individuals’ personality and socio-economic characteristics.

Read the paper

Stephen Hoskins, David W. Johnston, Johannes S. Kunz, Michael A. Shields, Kevin E. Staub

2024-05

Using a high-frequency panel survey, we examine the sensitivity of estimated self-reported well-being (SWB) dynamics to using monthly, quarterly, and yearly data. This is an important issue if SWB is to be used to evaluate policy. Results from autoregressive models that account for individual-level heterogeneity indicate that the estimated persistence using yearly data is near zero. However, estimated persistence from monthly and quarterly data is substantial. We estimate that persistence to shocks typically lasts around six months and has a net present value of 75–80 per cent of the contemporaneous effect. Estimates are similar for different domains of SWB.

Read the paper

Sofia Amaral-Garcia, Mattia Nardotto, Carol Propper, and Tommaso Valletti

2024-04

We examine the effect of internet diffusion on the uptake of an important public health intervention: the measles, mumps and rubella (MMR) vaccine. We study England between 2000 and 2011 when internet diffusion spread rapidly and there was a high profile medical article (falsely) linking the MMR vaccine to autism. OLS estimates suggest internet diffusion led to an increase in vaccination rates. This result is reversed after allowing for endogeneity of internet access. The effect of internet diffusion is sizable. A one standard deviation increase in internet penetration led to around a 20% decrease in vaccination rates. Localities characterised by higher proportions of high skilled individuals and lower deprivation levels had a larger response to internet diffusion. These findings are consistent with higher skilled and less deprived parents responding faster to false information that the vaccine could lead to autism.

Read the paper

Francis W. Graham, Sonja de New, Suzanne Nielsen and Dennis Petrie

2023-09

The proliferation of the high-dose prescription opioid product OxyContin has been identified as a major contributing factor to rising rates of opioid-involved harm throughout the early stages of the US opioid epidemic. Furthermore, after OxyContin was reformulated with abuse-deterrent properties in 2010, many people previously engaging in extramedical OxyContin use substituted to illicit substitutes, initiating a wave of heroin- and later synthetic opioid-involved deaths. Using event studies similar to those employed in previous OxyContin-related studies, we provide evidence that the OxyContin reformulation also induced substitution to another high-dose extended release (ER) prescription opioid product marketed under the brand name Opana ER. We show that the steady continued growth in prescription opioid-involved mortality after the OxyContin reformulation is nearly entirely explained by substitution from OxyContin to Opana ER. Furthermore, we show that when Opana ER itself was reformulated in February 2012, there was another wave of substitution to heroin previously attributed solely to OxyContin. Our estimates imply that if the Opana ER pathway were shut down at the time of the OxyContin reformulation, heroin-involved mortality from 2009 to 2016 in the US would have been as much as 33% lower, synthetic opioid-involved mortality as much as 38% lower, and total opioid-involved mortality as much as 44% lower. This study provides new evidence of the harms posed by high-dose prescription opioid products throughout the US opioid epidemic, as well as the unintended consequences of supply disruptions in the presence of both licit and illicit substitutes.

Read the paper

Nicole Black, Lachlan Deer, David W. Johnston, and Johannes S. Kunz

2023-08

Peer-to-peer support is often a critical component of mental health programs, but evidence on the effect of peer-based support programs at scale is limited. Using quasi-experimental methods, we examine whether a prominent peer-based support campaign, “R U OK? Day”, affects short-term mental health outcomes in Australia. Using variation in daily records and differences in the campaign’s intensity over nine years, we find no evidence that “R U OK? Day” reduces suicides and suicidal behaviours in the month after the campaign. However, we find positive effects on mental wellbeing, particularly among middle-aged males, with improved social support the likely mechanism. Our results provide evidence that peer support campaigns may be a practical, low-cost approach to improve population mental wellbeing.

Read the paper

Sundar Ponnusamy and Mohammad Abbas Hakeem

2023-07

We examine the association between ethnic inequality and various key health outcomes for a global set of developed and developing countries. Our results show that higher ethnic inequality is associated with a poor state of public health, such as higher child and maternal mortality, increased stillbirths and child stunting, and reduced life expectancy at birth. This set of effects is found to be predominant mainly in developing countries, and Sub-Saharan African countries. Results remain robust to the inclusion of various other measures of inequality, ethnic composition indices, geographic endowments, and other relevant controls. We argue that lower contraceptive usage and poor vaccination rates are potential mechanisms through which ethnic inequality affects health outcomes. Policies targeted at improving public health may need to focus more on these key intermediate channels in ethnic minority regions.

Read the paper

Nicole Black, Danusha Jayawardana, and Gawain Heckley

2023-06

Children’s time investments in various activities may be important for reducing socioeconomic status (SES) gaps in educational and mental health outcomes. Using time use diaries of Australian children aged 4-14, we find children from low SES backgrounds spend more time on digital media and less time on cognitively stimulating out-of-school activities, organised or for leisure. This difference contributes about 4% to the observed SES gap in numeracy skills. The contribution is larger for males, older age groups, and when the cumulative effect on learning is considered. No clear results are found for literacy skills and mental health outcomes.

Read the paper

Johannes Kunz and Anna Zhu

2023-05

We study the effect of reducing welfare assistance on migrants’ long-term integration in Australia. The policy postponed a migrant’s eligibility for benefits during their first two years in the country. It mainly affected mothers and was announced after their arrival. Using a regression discontinuity design and 21 years of administrative welfare data, we find significant reductions in welfare receipt, where the gap widened over time, and stabilized in the long run. Benefit receipt amounts reduced by 28%, and time-on-benefits by 19%, particularly in the unemployment and disability categories. We observe larger treatment effects for mothers from disadvantaged backgrounds.

Read the paper

Sofia Amaral-Garcia, Mattia Nardotto, Carol Propper and Tommaso Valletti

2023-04

We study the effect of internet diffusion on the uptake of the measles, mumps and rubella (MMR) vaccine for children in England between 2000 and 2011. OLS estimates suggest that internet diffusion led to an increase in vaccinations but this result is reversed once we instrument for internet access. We find that the effect of internet diffusion on vaccination rates is sizable: a change of one standard deviation in internet take up determined an approximately 20% decrease in vaccination rate. We also find that areas with a higher proportion of high skilled individuals and lower deprivation levels are those with a higher response to internet diffusion in terms of the reduction in MMR vaccination rates. These findings are consistent with higher skilled and less deprived parents responding faster to false information circulated at the time that the vaccine could lead to autism. Even though this information has been proven to be wrong, these parents were those absorbing it more.

Read the paper

Coen van de Kraats, Titus Galama and Maarten Lindeboom

2023-03

We provide evidence that the social norm (expectation) of work has a detrimental causal effect on the mental well-being of individuals not able to abide by it. Using SHARE data on men aged 50+ from 10 European countries, we identify the social norm of work effect in a difference-in-differences model that compares mental well-being scores of unemployed / disabled individuals (the treatment group) with those of employed / retired individuals (the control group) at varying levels of the fraction of retirees of comparable age. The initial mental well-being gap at age 50 is large, with unemployed / disabled men experiencing substantially lower levels of mental well-being, comparable to, e.g., the detriment of being widowed. Beyond age 50, the mental well-being of unemployed and disabled men improves as peers of comparable age retire, and full convergence occurs generally at an age that is slightly above the normal retirement age, when everyone has retired.

.

Read the paper

Hai-Anh Dang, Trong-Anh Trinh and Paolo Verme

2023-02

Hardly any evidence exists on the effects of mental illness on refugee labor outcomes. We offer the first study on this topic in the context of Australia, one of the host countries with the largest number of refugees per capita in the world. Analyzing the Building a New Life in Australia longitudinal survey, we exploit the variations in traumatic experiences of refugees interacted with post-resettlement time periods to causally identify the impacts of refugee mental health. We find that worse mental health, as measured by a one-standard-deviation increase in the Kessler mental health score, reduces the probability of employment by 14.1% and labor income by 26.8%. We also find some evidence of adverse impacts of refugees’ mental illness on their children’s mental health and education performance. These effects appear more pronounced for refugees that newly arrive or are without social networks, but they may be ameliorated with government support.

.

Read the paper

Johannes S. Kunz, Carol Propper, Kevin E. Staub and Rainer Winkelmann

2023-01

We examine variation in US hospital quality across ownership, chain membership, and market concentration. We use a new measure of quality derived from the penalties imposed on hospitals under the flagship Hospital Readmissions Reduction Program. We document a robust and sizable negative for-profit quality gap: for-profit hospitals are consistently of lower quality. We find that a substantial part of the gap is related to being located in less competitive markets. This reduction occurs most for hospitals that are part of large national chains. For such hospitals we find no quality gap in fully competitive markets.

.

Read the paper

Duncan Mortimer, Rohan Sweeney, Amelia Turagabeci, Sepesa Rasili

2026-04

Climate change is forcing difficult choices between in-place adaptation and relocation for Pacific Island communities, yet policy responses often rely on participatory planning frameworks that privilege louder voices or implicitly assume a consensus of preferences. We surveyed 476 adults across 25 at-risk Fijian villages using a discrete choice experiment to understand how individuals evaluate trade-offs between alternative future living arrangements, including location, services, housing, income opportunities, climate risk, and cultural connection. Our analysis identifies three distinct preference types — movers, stayers, and adapters — with sometimes conflicting priorities. While movers and adapters are generally willing to relocate to climate-resilient locations, stayers prefer to remain in their existing villages even in the absence of significant adaptation investment. These divergent preferences reveal relocation and in-place adaptation as spatially constrained and contested choices. Uncoordinated household-level decisions by movers and adapters risk redistributing rural populations across to urban centres and fragmenting communities. Preservation of connection to community and place may therefore require deliberate coordination and compromise at the community level, including the design of new climate-resilient settlements that accommodate the preferences of stayers. Recognising heterogeneous preferences and the limits of consensus-based participation is essential for designing community adaptation pathways that are socially, culturally, and spatially just and acceptable.

Read the paper

Lihini de Silva, David Johnston, and Sundar Ponnusamy

2026-03

Government funding for environmental disasters and climate adaptation can strongly influence community recovery and resilience-building. Yet given funding is often distributed via competitive schemes, inequities may arise if allocation is determined by factors such as cost effectiveness rather than need. Consequently, disadvantaged communities may receive inadequate support and be highly vulnerable to future disasters. We examine whether federal government community grants following the Australian 2019/20 Black Summer Bushfires were distributed equitably. Using detailed grant-level data including where the grant activity took place, the recipient organisation, and the amount awarded, we find that even after controlling for physical exposure to the fires, communities that are more vulnerable by demographics (e.g., more children, elderly people, non-working individuals, and First Nations people) receive less. Communities with larger ethnic minority populations also get less though this result is not as robust. Conversely, communities with greater built environment vulnerability (i.e., more remote) receive more. We demonstrate that lower funding for demographics and minority vulnerable communities manifest mostly via grants targeting economic and social outcomes whilst higher funding for built environment vulnerable communities is largely driven by infrastructure grants. Furthermore, inequities persist across organisation types including government institutions, which are expected to more carefully consider equity compared to non-government organisations. Finally, the observed inequities hold even across grants received by the same organisation. Altogether, our findings suggest a tension in competitive grant schemes between targeting need and funding projects that are more likely to be successful and cost effective, resulting in socially vulnerable communities receiving less.

Read the paper

Rohan Sweeney, Farzana Hossain, Jumriani Ansar, Indra Dwinata, Sitti Andriani Anwar, Arlyani Risal, Gang Chen, Michaela F. Prescott, S Fiona Barker, Karin Leder, Ansariadi Ansariadi, and David W. Johnston

2026-02

This study employs two discrete choice experiments (DCEs) conducted with two sample groups in Indonesia to investigate the informal settlement upgrading priorities of residents (sample 1) and explore how they align with public taxpayers’ preferences (sample 2). The first
DCE explores the relative importance placed upon common planning and public health priorities, such as water security, drainage, and diarrhoea in children, alongside local economic development priorities. The second DCE investigates the relative importance placed upon project implementation design considerations, including project completion time and community consultation. Our findings reveal that residents particularly prioritise improvements in water quality and economic development. While informal settlement upgrading interventions often prioritise improving water, sanitation and hygiene (WaSH) to reduce diarrhoea and other water-borne disease, our study highlights that residents also highly value economic empowerment, underscoring the need for integrated upgrading approaches that address both health and livelihood concerns. Taxpayer perspectives were well-aligned on upgrading outcome priorities, but diverged slightly on project implementation. Whereas residents prioritised minimising project duration and were less concerned with significant community consultation, taxpayers emphasised generating employment opportunities for residents within project designs. Both groups expressed an aversion to residents bearing full responsibility for resourcing ongoing operations and maintenance, preferring government or shared responsibility, highlighting the need for sustainable funding models. The study highlights the value of DCEs as a tool to support locally-led development, informing upgrading strategies that are more likely to be both politically feasible and successfully appropriated into urban livelihood practices of residents.

Read the paper

Nicole Black, Anthony Harris, David W. Johnston, and Trong-Anh Trinh

2026-01

This paper examines how the composition of financial resources relates to financial hardship and financial satisfaction among retired households. Using 23 waves of a nationally representative Australian panel, we distinguish between government pensions, private income, liquid financial assets, housing wealth, and debt. In fixed-effects models that exploit within-individual changes over time, liquid financial assets are strongly associated with lower financial hardship and higher financial satisfaction, while other forms of wealth show little independent association. We additionally examine responses to health shocks and find that liquid assets significantly attenuate their adverse financial effects, providing direct evidence of a buffering role. These results indicate that retirees with similar total wealth may experience different financial outcomes depending on portfolio composition, and suggest that adequacy assessments based on aggregate wealth or income replacement rates alone may overlook an important source of financial vulnerability.

Read the paper