Research
Working Papers
Child Health and Parental Responses to an Unconditional Cash Transfer at Birth, with John Lynch, Aurélie Meunier, Rhiannon Pilkington, and Stefanie Schurer.(Revise and Resubmit at the Review of Economics and Statistics).
Abstract
We estimate the impact on child health of the unanticipated introduction of the Australian Baby Bonus, a $3,000 one-off unconditional cash transfer at birth. With population-level administrative data from South Australia and a regression discontinuity design, we find that eligible infants experienced fewer hospital presentations in the first year of life for preventable, acute, and severe problems. In auxiliary analyses using nationally-representative data, we find that the windfall cash allowed parents to increase expenditures on food and groceries, decreased financial stress and hardship, and improved physical and mental health. We calculate that 34% of the pay-out were recouped within the first year through lower health care costs.Working Paper (October 2023) | IZA Discussion Paper (Sep 2021).
Slides NBER Summer Institute 2022 (July 2022).
(Revise and Resubmit at the Journal of the European Economics Association).
Abstract
We develop a method to detect local non-compliance with a random treatment to recover valid quasi-experiments from existing data. Our approach combines simulation-based methods and latent class modelling in an intuitive way, can detect and characterize non-compliant risk sets, and is computationally undemanding. To illustrate its usefulness, we use it to estimate ability peer effects in Taiwan, where we have unusually rich data and a national mandate to randomly assign students to classrooms within schools, but with partial non-compliance. After recovering a valid quasi-experiment, we first estimate ability peer effects in line with other studies. We then document precisely estimated null effects on 18 potential mechanisms, many of which have been hypothesized but never tested. Our application shows how to use our method to expand the causal evidence base using existing datasets.Latest Version (September 2024).
Older Version (“On the Mechanisms of Ability Peer Effects”, Dec 2020).
Media: The Conversation (Nov 2021) | ABC Radio Sydney Afternoons (15 Nov 2021).
(Revise and Resubmit at the Review of Economic Studies).
Abstract
The idea that students benefit from same-sex teachers has motivated many policies around the world aimed at reducing gender inequalities. However, we do not know the size or generalizability such same-sex teacher effects. We fill this gap by conducting a meta-analysis and our own study using data from 90 countries. Our meta-analysis summarizes the literature and highlights that studies are often underpowered and difficult to compare because they use different methods. Our multi-country study overcomes these shortcomings by providing many high-powered and comparable estimates. Those estimates are ideal for learning about the generalizability of an effect. Our results reveal an interesting pattern. In secondary education, same-sex teacher effects are positive for most countries and outcomes. In primary education, effects are more mixed. For example, we find negative same-sex teacher effects on test scores for about half the countries. Our paper showcases how we can use multi-country studies to learn about the generalizability of an effect.University of Zurich Working Paper No. 438 (June 2023, Updated May 2024) | Website with interactive results.
Abstract
We develop a multi-agent model of the education production function where investments of students, parents, and teachers are linked to the presence of minorities in the classroom. We then test the key implications of this model using rich survey data and a mandate to randomly assign students to classrooms. Consistent with our model, we show that exposure to minority peers decreases student effort, parental investments, and teacher engagement and it results in lower student test scores. Observables correlated with minority status explain less than a third of the reduced-form test score effect while over a third can be descriptively attributed to endogenous responses of the agents.CEPR Discussion Paper (April 2024).
Media: VoxEU Column.
Other: Best Paper Award at the 2023 Australian Labour Econometrics Workshop.
Abstract
This study pushes our understanding of research reliability by reproducing and replicating claims from 110 papers in leading economic and political science journals. The analysis involves computational reproducibility checks and robustness assessments. It reveals several patterns. First, we uncover a high rate of fully computationally reproducible results (over 85%). Second, excluding minor issues like missing packages or broken pathways, we uncover coding errors for about 25% of studies, with some studies containing multiple errors. Third, we test the robustness of the results to 5,511 re-analyses. We find a robustness reproducibility of about 70%. Robustness reproducibility rates are relatively higher for re-analyses that introduce new data and lower for re-analyses that change the sample or the definition of the dependent variable. Fourth, 52% of re-analysis effect size estimates are smaller than the original published estimates and the average statistical significance of a re-analysis is 77% of the original. Lastly, we rely on six teams of researchers working independently to answer eight additional research questions on the determinants of robustness reproducibility. Most teams find a negative relationship between replicators’ experience and reproducibility, while finding no relationship between reproducibility and the provision of intermediate or even raw data combined with the necessary cleaning codes.I4R Discussion Paper Series (April 2024).
Abstract
We use a rigorous three-stage many-analysts design to assess how different researcher decisions—specifically data cleaning, research design, and the interpretation of a policy question—affect the variation in estimated treatment effects. A total of 146 research teams each completed the same causal inference task three times each: first with few constraints, then using a shared research design, and finally with pre-cleaned data in addition to a specified design. We find that even when analyzing the same data, teams reach different conclusions. In the first stage, the interquartile range (IQR) of the reported policy effect was 3.1 percentage points, with substantial outliers. Surprisingly, the second stage, which restricted research design choices, exhibited slightly higher IQR (4.0 percentage points), largely attributable to imperfect adherence to the prescribed protocol. By contrast, the final stage, featuring standardized data cleaning, narrowed variation in estimated effects, achieving an IQR of 2.4 percentage points. Reported sample sizes also displayed significant convergence under more restrictive conditions, with the IQR dropping from 295,187 in the first stage to 29,144 in the second, and effectively zero by the third. Our findings underscore the critical importance of data cleaning in shaping applied microeconomic results and highlight avenues for future replication efforts.SSRN Working Paper (February 2025).
Abstract
We analyse the effects of a national student finance reform in the Netherlands, which replaced universal subsidies for higher education students by low-interest loans. We show that this reform had a large impact on education choices of secondary school students, lowering their enrolments in college-preparing tracks and increasing the share of students specializing in STEM subjects. The reform also affected the living arrangements of new college entrants. Our findings highlight that secondary school students respond to the modes of higher education financing well ahead of their graduation, and that financial aid uncertainty alone can deter many from pursuing higher education.IZA Discussion Paper (Aug 2021).
Abstract
Worldwide, single mothers are profoundly time and income constrained, making them heavily reliant on government transfers. We examine how welfare reforms that introduced mutual obligations affected the economic position of single mothers and the development of their children over the past two decades in Australia. Using nationally representative longitudinal data, we show that disposable incomes of single-mother households were significantly reduced relative to partnered mothers since the 2005 Welfare-to-Work Act came into effect in July 2006, a downward trend that was aggravated by the Global Financial Crisis and the 2013 suspension of grandfathered single parenting payment rules. The reform diminished parenting and family payments for single mothers, who compensated income loss by increasing reliance on disability pension payments, work hours, and child-care expenditures. We then use nationally representative cohort data to estimate the impact of single motherhood on child skill development, following children who entered primary school when their mothers were affected by the Welfare-to-Work reform. We find unadjusted single-motherhood gaps of 0.2 SD in cognitive and 0.3 SD in non-cognitive skills. Non-cognitive skill gaps are only partially explained by differences in observable characteristics, while cognitive skill gaps are fully explained by observable characteristics. Differences in disposable household income between single and partnered mother households explain over 50% of the observed cognitive ability gaps in childhood and 25% in late adolescence. In the presence of positive spillover effects, we propose that welfare payments to vulnerable families may function as a social investment rather than a sunk cost.Publications
Re-examining the relationship between patience, risk-taking, and human capital investment across countries (2024, Journal of Applied Econometrics, 1-9.), with Jan Feld, Nicolás Salamanca.Abstract
Hanushek et al. (2022) show that students in countries in which people are more patient and less risk-taking perform better in the Programme for International Student Assessment (PISA) test. In this paper, we probe the robustness of this study. Our narrow replication shows that the results are mostly robust to alternative model specifications. Our broad replication shows that the main results are robust to measuring student performance with data from The Trends in International Mathematics and Science Study (TIMSS) and The Progress in International Reading Literacy Study (PIRLS) instead of PISA.Selected Research in progress
Class Rank and Sibling Spillover EffectsAbstract
This paper studies how students’ relative rank affects their younger siblings’ academic performance. Using Dutch administrative data, I exploit within-school, across-cohort variation in student rank. Consistent with previous research, I find that higher-ranked students experience faster human capital accumulation. Yet, these rank effects generate negative spillovers on younger siblings: a one standard deviation increase in rank decreases younger sibling’s test scores by 2 percent of a standard deviation. These spillovers are stronger for same-sex siblings. I show that changes in parental investments and expectations drive part of these spillovers. In particular, I show that migrant parents increase investments in school quality and speaking Dutch at home. My findings highlight that the peer environment creates within-family human capital spillovers.Draft JMP version (31 August 2022).
College Information Provision to Disadvantaged Youth, with Gloria Chen and Siqi Pan.
Status: RCT in the field.