Structural breaks – including crises, jumps in volatility, and changes in regime – are per-
vasive in the economy. Detecting breaks can be a major empirical challenge: when and
where do they occur, how often, for how long, and of what magnitude? Not detecting
actual breaks can deleteriously affect economic analysis, forecasting, and policy. Auto-
mated model selection and impulse indicator saturation are two recent methodological
innovations that can help detect breaks. Their ability to detect breaks is relevant both
in-sample and out-of-sample, so these two tools offer improvements to existing method-
ology for empirical modeling, forecasting, and policy analysis. This paper illustrates and
generalizes these tools by re-analyzing the empirical model of seasonally unadjusted UK
narrow money demand in Ericsson, Hendry, and Tran (1994). Both tools demonstrate the
robustness of that model to a wide range of feasible alternatives. These tools also yield
statistical and economic improvements to that model, and so provide insights into the
practical justication of empirical evidence in macro-economics. Combined, these tools
permit computer-automated parsimonious detection of structural breaks.