Resumen:
This study examines the degree of time persistence in U.S. nuclear electricity net generation using innovative fractional integration and autoregressive models with monthly data from 1973:1 to 2011:10. The results indicate that nuclear electricity net generation is better explained in terms of a long memory model that incorporates persistence components and seasonality. The degree of integration is above 0.5 but significantly below 1.0, suggesting nonstationarity with mean reverting behavior. Policy implications are discussed as well.