When Information Fails: SMS Dissemination of Utility Outage Schedules and Firm Behavior in Urban Malawi

Utilities and governments in low-income countries increasingly use mobile channels to communicate with citizens, on the premise that access to official information is the binding constraint. We study a setting in which the constraint instead lies upstream, in the quality of the information itself. In a randomized pilot in Blantyre, Malawi, we relayed the electricity utility's published weekly load-shedding schedules to urban microenterprises via SMS, a channel that reaches firms at a small fraction of the cost of the newspapers and website where the schedules are posted. Using matched schedule data and high-frequency phone surveys on day-level operations and self-reported outages, we assess both the informativeness of published schedules and firms' behavioral responses. Published schedules are weak predictors of reported outages and systematically over-predict interruptions; about half of scheduled outage windows pass without a reported outage. Consistently, SMS dissemination does not lead firms to adjust operating hours around scheduled outages; instead, closures track realized outages reported by study businesses. Expanding digital dissemination creates little value when the underlying signal is uninformative. Last-mile communication investments require complementary institutional capacity to produce accurate, credible service information.