THE TWIN PHRASES "FULL EMPLOYMENT" and "reasonable price stability" usually appear at the beginning of any discussion of the goals of stabilization policy, whether it is a President's economic report, a journalist's financial column, or an economist's principles course. Unemployment and inflation are both "bads," which should be eliminated. Price stability is sufficiently important that if an increase in unemployment is required to eliminate inflation, either that extra unemployment must be tolerated (Nixon-Shultz pre-August-1971 "gradualism") or the inflation must be suppressed by partial or complete controls (many "Democratic econo- mists" and post-August Nixon). But full employment and price stability do not deserve equal primacy as stabilization goals. Most of the evils commonly associated with inflation occur only when the actual inflation rate deviates from that which is expected, that is, when the inflation is a surprise. The welfare costs of a fully anticipated inflation, although a popular subject in economics journals and graduate theory exams, are not widely discussed or understood elsewhere. In this paper I first discuss the costs of maintaining a steady inflation at a rate similar to the 5 percent annual average that has obtained in the United States since the beginning of 1968.