Full name: Sebastian Arechaga Garces
I am on the 2025–2026 job market.
My research spans labor economics, public economics, state capacity, and electoral preferences. I currently study policies designed to improve educational outcomes in underserved communities, with a particular focus on the role of teachers. Other projects include local government redistribution, workers' responses to incentives, and voters' preferences for policies.
Contact Information
Email: sebastian_arechaga@berkeley.edu
Phone: +1 (510) 833 8197
Job Market Paper
Performance-based incentives in the public sector often raise concerns that they may widen inequalities if rewards flow disproportionately to already advantaged institutions. This tension is especially salient in education, where teachers shape student achievement, but vulnerable schools struggle to attract and motivate talent. I study the Teacher Incentive Allotment (TIA) program, a statewide performance-based pay reform launched in Texas in 2019. The program provides substantial bonuses to highly effective teachers, with larger amounts for schools serving more vulnerable students.
Exploiting the staggered rollout of TIA across nearly 400 districts and 3,000 schools, I estimate its causal effects using an event-study design. TIA raised standardized test scores by 0.04 standard deviations on average, with the largest gains in vulnerable schools, reducing within-district inequality by 2.3–8.7%. Teacher turnover declined temporarily, creating a more experienced and stable workforce. Among movers, designated teachers were more likely to transfer to high-vulnerability schools, where bonuses were higher. Observable changes in teacher composition explain at most 14% of gains, implying that most improvements stem from higher instructional effectiveness. By disproportionately boosting achievement in disadvantaged schools, TIA narrowed educational gaps, showing that well-targeted incentives can enhance both efficiency and equity in education.
Work in Progress
What Drives Teachers' Electoral Engagement
I use data from teacher school rosters and records of contributions to parties to examine the following questions: Do teachers free-ride on others' contributions? Are they more responsive to the contributions of peers or neighbors? To what extent is teachers' contribution to parties affected by school districts?
Teachers' Contribution to Parties and School Performance in the Electoral Cycles
Using data from teacher school rosters, standardized test scores, and records of contributions to parties, I examine the patterns of contributions among teachers and its relationship with student academic performance. My analysis reveals that, in the post-COVID-19 period, teachers have made significantly more contributions than individuals in other occupations. Notably, teachers that contribute are often found in schools with lower test scores. The causal impact of teacher electoral participation on student achievement remains an area of ongoing research.
Government Size and Ideology. The Case of California Local Bond Measures
In California, local governments must obtain voter approval through ballot measures to issue general obligation bonds backed by property taxes, particularly for school and infrastructure projects. These ballots specify how the funds will be used, providing a clear link between voters’ political preferences and their support for different sizes of government. Using precinct-level data from school finance ballot measures in the largest California counties, I document two main findings. First, precincts with more conservative electorates are less likely to approve bonds and property tax increases. Second, these same precincts exhibit higher voter turnout in such elections.
In ongoing work, I develop a model that rationalizes these patterns by incorporating uncertainty about the benefits of proposed spending and the relative certainty of its fiscal costs. The model aims to explain how ideological preferences and perceived risk shape voter behavior toward local government size and public investment.
