Mehrdad Baratian

Mehrdad Baratian

Ph.D. Candidate in Economics

Tulane University · New Orleans, LA

About

I am a Ph.D. candidate in economics at Tulane University, working in public economics, health economics, applied microeconomics, and the economics of innovation. Much of my work asks how public programs and large shocks change behavior in ways that only surface years later.

I am drawn to questions where the key variable has never been measured. The part I enjoy most is pulling evidence out of unstructured data, the text and records that never arrive as tidy tables, and turning it into measures I can use to study welfare programs and the health and economic outcomes they shape.

My job market paper studies the U.S. safety net. Using the Panel Study of Income Dynamics, I ask how the 1990s shift to “work-first” welfare changed the time environment of early childhood for a generation of children, and how that exposure surfaces in their adult outcomes decades later. Alongside it, one project asks how the generosity of SNAP benefits supports mental health when households are under strain, and another uses COVID-19 to trace how a large shock redirects scientific research effort across fields, pairing an NLP measurement of roughly four million PubMed papers with event-study methods.

I lean on quasi-experimental designs and machine learning, and I care about measurement as much as identification.

Research

The Effects of Early Childhood Exposure to Welfare Reform on Adult Health and Employment

Job Market Paper

Latest version (PDF)

In models of child development, a child’s later outcomes depend partly on family income and parental time. Welfare rules that require work trade one for the other, so theory does not say whether children gain or lose. I study the 1990s welfare reforms that imposed such rules, following young children into adulthood in the Panel Study of Income Dynamics. I compare children born in the same year across states whose rules differed in strictness, by their families’ likelihood of receiving welfare. Among children the rules likely reached, those with more exposure to stricter rules before age five had better general and mental health as adults. But their adult employment did not change, and the estimates for their hours and earnings are imprecise. On net, my findings suggest that these children gained as adults from the trade-off between family income and parental time that the rules imposed.

Shocks, Reallocation, and the Direction of Innovation: The Case of COVID-19

Working Paper

Working paper · draft available on request

When a shock raises the return to one line of research, researchers can change fields or absorb the new questions into the work they already do. This paper measures which margin carried the adjustment of biomedical research to COVID-19. A classifier built with BioBERT, a biomedical language model, assigns every PubMed publication from 2018 through 2022, an estimation sample of about four million papers, a probability of belonging to each of thirty-one research fields. Averaging these probabilities within a quarter gives each field’s share of attention, and I estimate the change in each share at the first quarter of 2020.

Read full abstract

Every one of the thirty-one field shares broke at the onset, and all survive a false discovery correction. The COVID field absorbed about three percentage points of total attention. Sixteen fields gained and fifteen lost, and the pattern did not follow proximity to coronavirus science. The adjustment ran within the fields instead. Researchers raised the coronavirus content of papers inside their home fields by 3.2 percentage points, and by 5.8 when papers are weighted by journal quality, and the rise appears at every tier of venue quality. This combination is what a model predicts when changing fields is costly and blending shock content into ongoing work is cheap. For a funder trying to redirect science in a crisis, the responsive margin is within fields rather than between them.

SNAP Generosity and Mental Health

Submitted

Submitted · draft available on request

Methods. Difference-in-differences; dose–response difference-in-differences; XGBoost machine learning for data imputation.

Finding. SNAP allotments support mental health during economic hardship, with effects emerging beyond a minimum benefit threshold.

Teaching

Research & Teaching Assistant, Tulane University

2022–present

  • Econometrics II (Ph.D.)
  • Machine Learning and Econometrics
  • Intermediate Microeconomics
  • Health Economics
  • Introduction to Macroeconomics
  • Introduction to Microeconomics

Teaching Assistant — Strategic Management, Tehran Polytechnic

Spring 2018

Curriculum Vitae

Download CV (PDF) Updated July 2026

Contact

Reach me

Mbaratian@tulane.edu

Department of Economics
Tulane University
New Orleans, LA

References

Kevin Callison — Associate Professor, Health Policy & Management, Tulane (Dissertation Chair)
Katy Bergstrom — Assistant Professor, Economics, Tulane
Jonathan Holmes — Assistant Professor, Economics, University of Ottawa