Working Papers
Beyond Expectations: Does Belief Uncertainty Matter for Parental Investment Decisions? (Job Market Paper)
A growing body of research measures and subsequently accounts for how parents perceive the average returns to investment in the child development process and how this uncertainty impacts their actual investment in children. This paper moves this research beyond the mean and is the first to allow for uncertainty in parental beliefs. I first develop and collect an elicitation procedure of parental subjective belief distributions concerning the technology of skill production and their subjective investment costs that is guided by a model of parental investment. Second, I extend existing measurement error correction methods that are required with belief estimation. I show that there is a significant amount of heterogeneity in parental beliefs and ignoring this can overstate the importance of socio-economic background in explaining parental investment. Parents hold downward-biased mean beliefs about the returns to investment. Moreover, parents who have higher mean beliefs also hold lower levels of uncertainty about their beliefs. Both mean beliefs and uncertainty correlate with actual investment measures. Finally, I estimate a model of parental investment with reference-dependent preferences and show that even though parents hold low mean beliefs, they have a strong incentive to invest if their child is at risk of experiencing a developmental delay.
One Says Goodbye, Another Says Hello: Turnover and Compensation in the Early Care and Education Sector - with Flavio Cunha; [NBER Working Paper 31869]; Submitted
The quality of the early environment children experience profoundly influences their human capital development. We investigate retention and compensation of the Early Care and Education workforce by merging datasets from three different government agencies in Texas. We construct a longitudinal panel of individuals born between 1980 and 1989. We study their educational background and their lifecycle labor market dynamics. Turnover in the ECE sector is 37% higher and earnings are 12% to 20% below those in the sectors that employ similar types of workers. We uncover steep educational and racial differences in turnover rates and earnings. We estimate a dynamic discrete choice model of schooling and work decisions for a sample of potential entrants of the ECE sector. We estimate an extensive labor supply elasticity of 2 but a turnover elasticity of -0.5. Our model predicts that entry and retention bonuses would have a small positive impact on recruitment and retention in the ECE industry.
An Equilibrium Model of School Teacher Wages and School Financing, with Aloisio Araujo, Gabriel Sallum [Draft available upon request]
School funding equalization is a common policy in the United States that redistributes revenue from property-rich school districts to property-poor ones. This paper examines the impact of the Texas school funding equalizatio policy, known as Recapture, on teacher wages and skill composition using a structural equilibrium model. We first use a policy change in 2019 that changed school funding formulas in Texas to understand how school districts allocate new revenue between payroll and other administrative functions. We find that about 20% of new revenue is put into payroll mainly via wage increases. The other 80% goes towards bureaucracy functions like consulting, contracting and legal services. We then build and estimate an equilibrium model of teacher and non-teacher labor supply and school district budget allocation informed by our reduced form evidence. Counterfactual simulations reveal that eliminating the recapture policy allows wealthier districts to offer higher wages and attract more teachers but may lower average teacher quality due to an influx of less-skilled individuals. Less affluent districts face budget reductions and reduced teaching resources but could experience an increase in average teaching quality. Our findings highlight important trade-offs in policy design regarding funding redistribution and educational equity.
Work in Progress
Mental Health and Human Capital Formation in Children; Data cleaning in progress
This project uses a Brazilian longitudinal dataset that follows families identified as at risk of depression over ten years. The data include detailed measures of parental and child mental health, genetics, socioeconomic background, and child outcomes. I study how depression in parents and children affects the process of human capital formation by estimating a skill production function that incorporates mental health as an input. I also examine whether income support programs, such as cash transfers, influence the relationship between mental health and child development.
Using Structural Models to Devise Policies for the Fertility Challenges Across Countries, with Steven Lehrer and Jin Young Yoon; Data assembly in progress
Across countries, fertility rates have declined significantly, with many developing countries now below the replacement level of 2.1 births. Reduced form models exploring differences across cohorts in high-income countries conclude that period-based explanations focused on short-term changes in income or prices cannot explain the widespread decline. We extend this research by considering estimation of dynamic structural models using cohorts from seven high income countries to try and shed new policy insights. By identifying and estimating the ‘deep’ parameters that describe preferences (technologies) and constraints of the decision-making process, we can decompose the heterogeneity in parameters across countries and cohort. Last, we will use our estimates to provide counterfactual predictions and shed light on how policymakers in different countries should develop policies to promote fertility.
Changed Parenting Practices: Between Increased Input Into Childcare and Worsened Outcomes for Children, with Weili Ding, Michael Kottelenberg, Steven Lehrer; Analysis in progress
Using novel data on the allocation of time to different activities to specific children in the same family we use triple difference estimators to examine how childcare policy influences these parental decisions. Consistent with prior research, we find that parents of children who gain access to subsidized childcare decrease their educational investments at home, with gender differences in the effect on time spent playing with the child. However, we additionally document that there is a statistically significant increase in parental investments in the older child. These increases are not driven by changes in maternal labor supply and shows how policies can shift the allocation of time on leisure activities and investments across siblings.
Policy Briefs
Identifying Impact in the Context of Recovery: TWC Child Care Business Coaching Program, Texas Policy Lab (2022)
Wage Supplementation Policies for Childcare Workers in Texas, Texas Policy Lab (2023)