For both IP and the CPI, the diffusion index (DI) forecasts were compared to benchmark forecasts from an autoregressive model and multivariate regression-based forecasts using various leading indicators. For the CPI, as an additional comparison forecasts were also computed using models based on the Phillips curve. These various forecasting models are now described in turn.

Diffusion Index forecasts. Diffusion index forecasts were computed as outlined in section 2. Two sets of variables were used: a balanced panel of 170 monthly macroeconomic time series, 1960:1 – 1997:9, and an unbalanced panel in which these 170 series were augmented by 54 monthly series which are available for only part of this period, so that the total number of series in the unbalanced panel is 224.

The series were selected judgmentally to represent 14 main categories of macroeconomic time series: real output and income; employment and hours; real retail, manufacturing and trade sales; consumption; housing starts and sales; real inventories and inventory-sales ratios; orders and unfilled orders; stock prices; exchange rates; interest rates; money and credit quantity aggregates; price indexes; average hourly earnings; and miscellaneous. The full list is given in Appendix B. This list is similar to lists which we have used elsewhere (Stock and Watson [1996, 1998]). These series were taken from a somewhat longer list, which was scanned visually to eliminate gross problems such as series redefinitions. However no further pruning of this list was performed based on forecast performance measures. The series were taken from the February 1998 release of the DRI/McGraw Hill Basic Economics database (formerly Citibase). In general these series represent the fully revised historical series available as of February 1998.