Predicting the Local Impacts of Energy Development: A Critical Guide to Forecasting Methods and Models

Models forecasting second-order impacts from energy development vary in their methodology, assumptions, and quality. As a rough dichotomy, they either simulate community development over time or combine various submodels providing community "snapshots" at selected points in time. Using one or more methods -- input/output models, gravity models, econometric models, cohort-survival models, or coefficient models -- they estimate energy-development-stimulated employment, population, public and private service, and government revenues and expenditures at some future time (ranging from annual to "average year" predictions) and for different governmental jurisdictions (municipal, county, state, etc.). Underlying assumptions often conflict, reflecting their different sources -- historical data, comparative data, surveys, and judgments about future conditions. Model quality, measured by special features, tests, exportability and usefulness to policy-makers, reveals careful and thorough work in some cases and hurried operations with insufficient in-depth analysis in others.