The distributional effects of power outages on regional economies: evidence from dynamic panel data and machine learning models

Description

When the power goes out, factories stall and shops close, but the losses may not fall evenly across a state. The authors link outage records to state-level output and inequality in the United States. Using a dynamic panel method built for such data, they relate two standard reliability measures, outage duration and frequency, to real GDP and the Gini index of income inequality. Machine learning models check whether the reliability measures carry real predictive weight.

A one percent rise in interruptions trims real GDP by somewhere between 0.07 and 3.7 percent, depending on the measure, and nudges inequality up modestly. The long-run bill for frequent outages passes two trillion dollars in lost output. The authors argue that reliability is not just an engineering target: it shapes regional prosperity and fairness. They urge grid investment as economic policy, not only as maintenance.

Catalog record

Source: https://europepmc.org/articles/PMC13135604