Research Projects

I have established a research program on the economics of human development. The various components of this program zoom in on the determinants of skill and health capital over the lifecourse, with a specific focus on non-cognitive skills and preferences. Over the last three years I have established collaborations with the medical sciences and epidemiology, and am involved in various data-linkage projects. I use this infrastructure to study human development of at-risk groups:
  1. Impact of public policy and interventions on human development
  2. Non-cognitive skills, education, parenting, and performance of disadvantaged children
  3. Methodological contributions to non-cognitive skill measurement
The ultimate aim of my research is to improve a country's productivity by scoping talent from disadvantaged families.

Below you find a list of my ongoing research projects. Please visit my publications site for completed research projects.


1. Impact of public policy and interventions on human development

My research is mainly funded through an Early Career Fellowship granted by the Australian Research Council (A$380,000), a Partnership Project (A$950,000), and a Centre for Research Excellence (A$2,500,000), both funded through the National Health and Medical Research Council. My empirical research is conducted with linked administrative data to follow the life of children from birth to late adolescence, and through linked administrative and survey data (that I have collected myself or which is accessible through the Longitudinal Survey of Australian Children). The following research projects are (almost) completed - Papers 1.-5. will be presented at a special session on Economics of Human Development at the First Inaugural Conference of the Australian and Asian Society of Labour Economics (AASLE), 7-9 December 2017:

1. Early life health investment and childhood development: Evidence from special care nursery assignment in Australia's Northern Territory (with Kevin Schnepel) 

2. Do welfare restrictions improve child health? Estimating the impact of income management in the Northern Territory (with Mary-Alice Doyle and Sven Silburn)

3. The Effect of Quarantining Welfare on School Attendance in Indigenous Communities (with Deborah Cobb-Clark, Nathan Kettlewell, and Sven Silburn)

4. Baby Bonuses and Early-Life Health Outcomes: Using Regression Discontinuity Design to Evaluate the Causal Impact of an Unconditional Cash Transfer (with John Lynch, Aurelie Meunier, and Rhiannon Pilkington)

5. Bonus skills: Examining the effect of an Australian unconditional cash transfer on child development. LCC Working Paper nr 2017-04 (with Jason Gaitz) [LINK]


In another important project I collaborate with Agnieszka Tymula on an intervention in which we study the impact of a lifestyle intervention on non-cognitive skills and economic preferences of at-risk patients. We collaborated with the Boden Institute of Obesity (University of Sydney), to explore the extent to which skills and preferences can be changed for an at-risk population when undergoing an intensive lifestyle treatment. We elicit risk and time preferences through experiments, and survey-based non-cognitive skills and vignettes at baseline, month 6 and month 12 follow-up. The intervention is currently underway, for more information see HERE. This research is funded through the ARC Life Course Centre Seed Funding (A$40,000) and a University of Sydney Faculty Research Grant (A$30,000).

6. Economic preferences, non-cognitive skills and health outcomes: Protocol for a beahvioural economics component in a large-scale weight loss intervention with at-risk adults.  (With Rosemary Elkins, Agnieszka Tymula, Nicholas Fuller, and Ian Caterson)

7. Obesity and Economic Preferences (With Chiara Pastore and Agnieszka Tymula)

8. Effectiveness of a lifestyle intervention to change preferences (with Agnieszka Tymula)


2. Non-cognitive skills, education, parenting, and disadvantage

In a series of empirical (and one theoretical) research projects, I demonstrate the role that parents and disadvantage play in shaping human development. Some are summarised here:

9. Non-cognitive skill development over the lifecourse: The role of father involvement in a child's education (with Rosemary Elkins) - submitted to Demography

Abstract: Using longitudinal British cohort data, we examine the role that fathers play in the maturation dynamics of internal locus of control (LOC) - a widely studied, beneficial non-cognitive skill -- from childhood into middle age. We use a machine-learning algorithm to first describe the most common LOC maturation pathways. Although three quarters of individuals exhibited high levels of internality throughout adulthood, their childhood internality scores show a wide spread: over 50% of individuals moved from low or average levels of childhood internality to high adulthood levels, and 25% experienced high levels of internality throughout the lifecourse. Estimating a standard skill production function, we find that father's, but not mother's, interest in their child's education at age 10 as assessed by the teacher predicts internality in middle age for female and socioeconomically disadvantaged children. Father's educational interest significantly increases the probability of lifelong internality by 20%, and protects against lifelong externality. Parental engagement in children's education is a malleable factor that has been the focus of many interventions, and thus is a promising target for public policy.

10. Do parenting styles affect the effectiveness of educational investments? Theory and empirical evidence

11. Child maltreatment and life-time health outcomes: Evidence from the National Child Development Study (with Tara Hariharan)

12. What are the pathways that link child maltreatment with life-time economic outcomes? (with Kristian Trajkovski)

Abstract: Many studies have examined the link between Adverse Childhood Experiences (ACE), these the experience of neglect, abuse, family instability, or illness in the family among others, and later-life health outcomes. However, little is known about the link between ACE and life-time economic outcomes. In this study, we will explore the relationship between ACE and economic outcomes – earnings, welfare dependency and poverty – attained by age 55. To understand the mechanisms through which ACE is linked with later-life economic outcomes, we decompose the estimated relationship into the potential mediating factors such as cognitive skills, non-cognitive skills, educational attainment, health capital, and family formation decisions in young adulthood. The empirical analysis is conducted with data from the National Child Development Study (NCDS), a birth cohort of children born within one week in 1958 in the UK and followed until today. The data allow us to measure ACE prospectively, deriving information from age 7-11 household surveys and teacher reports of neglect, and to control for a series of potential confounding factors that were recorded before the exposure to ACE, in particular  (e.g. household socioeconomic background). Our estimation results suggest that each additional ACE is associated with an earnings penalty of 7.3 percent by age 55, with a 53.1 percent higher probability of being welfare dependent, and a 34 percent higher probability of subjective poverty. These associations are mainly driven by cohort members’ experience of neglect, an assessment referring to malnutrition and dirty appearance recorded by the teacher when the cohort member was 7 or 11 years old. Decomposing observed differences in economic outcomes between those who were exposed to multiple ACE, and those who were not exposed, into the contribution of differences in observable and unobservable characteristics, we find that 45 percent of the difference in earnings are explained by differences in human capital accumulation by age 33. Observed differences in welfare dependence and subjective poverty are explained to a large proportion by differences in unobservable characteristics. The findings suggest that early-life exposure to ACE, especially when it includes parental neglect, is associated with significant economic penalties pre-retirement age, and such penalties are mediated through penalties in human and health capital attainment. 


As part of my ARC Fellowship, I have embarked on various projects that seek to quantify - what I call - exceptional upward mobility: Exceptionally upward mobile individuals are those who grew up in disadvantaged families and neighbourhoods but who achieved - against all odds - unexpected socioeconomic outcomes in adulthood. Some examples are for instance children whose father worked in a manual-unskilled profession but who became lawyers or university professors. The main idea of this line of research is to understand how much talent there is in disadvantaged families and whether this `talent' can be lost if left un-nurtured . The following projects are currently in preparation:

13. Quantifying the empirical distribution of exceptional upward and downward occupational mobility in three countries (Australia, Germany, Britain). With Angus Wheeler.

14. Early signs of exceptional upward mobility - We measure how many children from disadvantaged backgrounds are achieving high levels of cognitive and socio-emotional functioning (With Francisco Azpitarte). 

15. How well are students who are the first in their families to go to university prepared for university life? Evidence from first-years students at a university in NSW? (With Rachael Gibson, Rebecca Edwards and Colm Harmon) 


3. Methodological contributions

16. Survey item-response behavior as a proxy for unobserved ability: Theory and application (with Sonja Kassenboehmer) submitted to Journal of Political Economy

Abstract: We develop and test an economic model of the cognitive and non-cognitive foundations of survey item-response behavior. We show that a summary measure of response behaviour -- the survey item-response rate (SIRR) -- varies with cognitive and less so with non-cognitive abilities, has a strong individual fixed component and is predictive of economic outcomes. We demonstrate the usefulness of SIRR as a proxy for cognitive ability to reduce omitted-variable biases in estimated wage returns. We derive both necessary and sufficient conditions under which the use of a proxy-variable approach reduces omitted-variable biases, providing a new guideline for informed risk evaluation.

17. Using vignettes to control for scaling bias in personality assessment: We are currently collecting data on first-year students from the University of Sydney on their personality, how they rate the personality of others, and their family background. We developed vignettes - i.e. hypothetical persons - whose personality the survey participants are asked to rate in a similar way as they are asked to rate their own personality. We use these vignettes to adjust for the high potential of reporting heterogeneity in personality assessments that rely on Likert scales. The survey is currently in the field. Visit our webpage HERE (With Rebecca Edwards and Colm Harmon).