This thesis looks at how to use special statistical models to understand large macroceonomic datasets. It gives economists new methods to decompose major changes in the economy, produce better forecasts even during unstable times, and estimate the effects of economic shocks by using a mix of big data and additional external information.