Business Behavior in the Face of Severe Weather: Studying the Effects of Deterministic and Probabilistic Warning Systems Weather, Climate, and Society

Tornadoes are nature’s most violent storm and annually cause billions of dollars in damage along with the threat of fatalities and injuries. To improve tornado warnings, the National Weather Service is considering a change from a deterministic to a probabilistic paradigm. While studies have been conducted on how individual behavior may change with the new warnings, no study of which we are aware has considered the effect this change may have on businesses. This project is a response to the Weather Research and Forecasting Innovation Act of 2017, House of Representatives (H.R.) bill 353, which calls for the use of social and behavioral science to study and improve storm warning systems. The goal is to discuss business response to probabilistic tornado warnings through descriptive and regression-based statistics using a survey administered to businesses in north Texas. Prior to release, the survey was vetted by a focus group composed of businesses in Grayson County, Texas, who assisted in the creation of a behavior ranking scale. The scale ranked behaviors from low to high effort. Responses allowed for determining whether the business reacted to the warning in a passive or active manner. Returned surveys came from large and small businesses in north Texas and represent a wide variety of industries. Regression analysis explores which variables have the greatest influence on the behavior of businesses and show that, beyond increases in probability from the probabilistic warnings, trust in the warning provides the most significant change to behavior. Grant no. NA16OAR4320115