Forecasting • Causal Inference • Python / R / SQL / Stata
Working papers and research projects in applied economics, causal inference, and energy policy. Full methodology, code, and documentation are available for each paper.
This paper examines how much of ERCOT's 27 percent electricity demand growth between 2019 and 2025 is associated with data center investment, using generation interconnection queue filings as a proxy treatment variable. A four-layer identification strategy — panel regression, synthetic control, difference-in-differences with low-exposure controls, and narrative validation — produces a consistent estimated range across methods. The headline result is a 34.8 index-point divergence in ERCOT minimum hourly demand above a synthetic counterfactual built from balancing authorities with low data center exposure. A three-model forecasting pipeline extends scenario projections through 2027. The paper documents the absence of public load-side queue data as an explicit policy finding.
University of Maryland — Preliminary Draft — March 2026
SSRN (pending) • GitHub • Interactive Dashboard
March 2026 • Synthetic Control • DiD • Panel Regression • ARIMA • XGBoost • Python • Stata