This is an example of an "observable index model" from Sargent & Sims(1977), "Business cycle modeling without pretending to have too much a priori economic theory," in New Methods in Business Cycle Research: Proceedings from a Conference, Federal Reserve Bank of Minneapolis. This is famously the one paper co-written by the 2011 Nobel laureates.
This paper is heavily cited in the dynamic factor model literature, but neither of the techniques actually used in that paper: the observable index model (Sims) or the unobservable (frequency domain) index model (Sargent) form the basis for what is done nowadays. The observable index model is a reduced form VAR; Sims' hope was that it was restricted enough to provide decent forecasts when a full VAR couldn't because of overparameterization. For that purpose, it was relatively quickly supplanted by BVAR's, which could be computed much faster. There's nothing inherently wrong with the model; it just was proposed at a time that it was too costly to use.
The data set is from Bernanke, Boivin & Eliasz (2005), "Measuring the Effects of Monetary Policy: A Factor-augmented Vector Autoregressive (FAVAR) Approach," The Quarterly Journal of Economics, vol. 120(1), pages 387-422. The model is applied to 15 Industrial Production series.