Working Papers

Previous version's title "Spillover Effect of Prescription Drug Monitoring Program Integration on Prescription Opioids Consumption"

Presentations: CIST 2019; ICIS 2019; SCECR 2020

Many states in the U.S. have promoted integrating Prescription Drug Monitoring Programs (PDMPs) into local electronic health records (EHRs) to combat the opioid crisis. We document that such within-state PDMP-EHR integration in a neighboring state could increase the quantity of consumed prescription opioids in counties of a focal state.  This undesired spillover effect could be mitigated when the two states have interstate PDMP data sharing.

* Corresponding author

Presentations: CODE@MIT 2022; ESSEC-HEC-WBS IS Workshop 2022 

We propose a novel experimental methodology to estimate the economic value of algorithmic interventions in the setting of online advertising, which is a two-sided online marketplace. Our method prevents various sources of interference and greatly reduces the variance of estimates, compared to conventional A/B testing approaches. 

Online consumers adopting privacy-preserving techniques might simultaneously enjoy an increased intrinsic value of privacy (e.g., stronger perception of privacy control regardless of economic outcomes) and a decreased extrinsic value of privacy (e.g., less personalized recommendations). We empirically separate the intrinsic and extrinsic components of consumer privacy, and link them to consumer behavior, using a unique clickstream dataset provided by a reputable news publisher.

Working in Progress

Presented at INFORMS Annual Meeting 2021

Research Collaboration with Industrial Partners