A monetary authority with the primary objective of price stability has to distinguish between temporary price shocks and persistent shocks to the rate of inflation. A measure of underlying inflation, therefore, has an important role to play as a guideline for monetary policy. In this paper, a measure of underlying inflation in the United States is obtained using a structural vector autoregressive (SVAR) methodology. The assumption that movements in measured inflation are the result of (a) one-time shocks to prices arising from supply-side developments and (b) persistent shocks to the inflation rate arising from demand-side developments provides a set of long-run restrictions to identify the structural innovations to the consumer price inflation rate.

The model is estimated with monthly data and includes consumer prices (CPI), capacity utilization (CAPUT), producer prices of finished consumer goods (PPI), and import prices (IMP). The evidence reported in this paper suggests that measured inflation in the United States was below its underlying trend rate in 1994 and 1995, a period when inflationary pressures remained subdued, despite above-potential growth and labour market tightness. The evidence also supports the view that temporary factors have helped to contain inflationary tendencies in recent years. Past shocks should exert some further downward pressure on the inflation rate. Moreover, we find that the tightening in 1994-95 coincided with an upward trend in the underlying inflation rate, while measured inflation was still trending downward. The finding supports the view that the Federal Reserve reacts to movements in the underlying trend inflation rate.