The parts are more than the whole: separating goods and services to predict core inflation

Economists have not been altogether successful in their efforts to forecast “core” inflation—an inflation measure that typically excludes volatile food and energy prices. One possible explanation is that the models used to make these forecasts fail to distinguish the forces influencing price changes in core services from those affecting price changes in core goods. While core services inflation depends on long-run inflation expectations and the degree of slack in the labor market, core goods inflation depends on short-run inflation expectations and import prices. By using a composite model that combines these different sets of explanatory variables, the authors of this study are able to improve upon the inflation forecasts produced by a standard model.