PHD CANDIDATE, KELLOGG MECS

Contact Information

Kellogg School of Management
Northwestern University
2211 Campus Drive
Evanston, IL 60208

Phone: 949-243-7184

hossein.alidaee@kellogg.northwestern.edu

 

Education

Ph.D., Managerial Economics, Northwestern University, 2023 (expected)
MA, Economics, Northwestern University, 2017
BA, Mathematics (Honors) and Economics, Macalester College, 2013

Primary Fields of Specialization

Development Economics, Behavioral Economics

Secondary Fields of Specialization

Economics of Networks

Curriculum Vitae

Download Vita (PDF)

Job Market Paper

“How Uncertainty About Heterogeneity Impacts Technology Adoption”

Individuals can learn about new technologies through peers or through more official sources. Peers’ information is often based on only a handful of experiences. By contrast, official sources, such as the government, back their information with rigorous testing. In my setting of agricultural technology adoption, government recommendations are no more effective at inducing adoption than peers. This implies that data from peers is more effective per datum. I propose that this arises because returns to technology adoption are heterogeneous based on context and individuals face uncertainty about the context where government testing took place. I confirm this mechanism using a lab-in-the-field experiment with 1,600 small and marginal farmers in Odissa, India. I also demonstrate that both survey data and results from a broad set of recent field experiments on agricultural extension are consistent with my mechanism that farmers place greater value on information with less context uncertainty.

Fieldwork Completed October 7th. 

Coverage: World Bank Development Impact Blog

Other Research Papers

“Recovering Network Structure from Aggregated Relational Data using Penalized Regression” with Eric Auerbach and Michael P Leung

Social network data can be expensive to collect. Breza et al. (2020) propose aggregated relational data (ARD) as a low-cost substitute that can be used to recover the structure of a latent social network when it is generated by a specific parametric random effects model. Our main observation is that many economic network formation models produce networks that are effectively low-rank. As a consequence, network recovery from ARD is generally possible without parametric assumptions using a nuclear-norm penalized regression. We demonstrate how to implement this method and provide finite-sample bounds on the mean squared error of the resulting estimator for the distribution of network links. Computation takes seconds for samples with hundreds of observations.

Example Python code and R package

CoverageWorld Bank Development Impact Blog

Works In Progress

“The Impact Of A Graduation Program On COVID-19 Resilience” with Jessica Goldberg, Dean Karlan, Mushfiq Mobarak, and Chris Udry

Graduation programs are multi-faceted programs to lift the ultra-poor out of poverty, consisting of a grant of a productive asset, training for how to convert the asset into an enterprise, life coaching, and the provision of a savings account. We evaluate the NGO BRAC’s latest rendition of the graduation program in Bangladesh for their 2020 cohort. This evaluation was impacted by the unexpected shock of COVID-19 partway through the program. Consequently, we focus on how the graduation program impacted participants’ economic resilience to COVID-19 and accompanying policies designed to mitigate its spread. Information about the original experimental design available upon request.

Endline Completed June 2022

“Social (Mis)learning: Evidence From Bangladesh” with Zack Barnett-Howell

Though social learning is an important channel for information about new technologies, it can sub-optimally reduce adoption by spreading inaccurate information from noisy observations. We conduct a randomized control trial in Bangladesh for an alert system that enables potato farmers to take precautionary measures against a crop disease. The alerts are personalized using each farmer’s self-reported anticipated sowing date. Because the reported sowing date was not always followed, we experience natural variation across villages in the number of farmers receiving accurate information about when to apply the recommended interventions. Using this natural variation, we study how inaccurate information about the technology’s efficacy, caused by farming reporting error, spreads through villages and impacts adoption of the alert system.

Endline Completed April 2020

“Matrix IV” with Eric Auerbach and Isaac Loh

Abstract available upon request

References

Prof. Chris Udry (Committee Co-Chair)
Prof. Lori Beaman (Committee Co-Chair)
Prof. Ben Golub