TO THIS POINT we have examined only anticipated increases in the fiat money stock. In this chapter we examine the effects of monetary surprises – unanticipated fluctuations in the fiat money stock – on output, in particular. As we do so, we also study the more general question of how data correlations resulting from policy surprises may mislead naive policy makers about the effects of the sustained implementation of their policies.

The Data

The Phillips Curve

In 1958, A.W. Phillips discovered a significant statistical link between inflation and unemployment for the United Kingdom over a century. Subsequent work uncovered the same correlation for many other economies. Although it was not understood why such a correlation existed, this discovery excited many in the economics profession by suggesting that there may be an exploitable trade-off between inflation and unemployment – that by increasing inflation, the government might achieve lower unemployment and greater output. The apparent inverse relationship between inflation and unemployment rates that existed in the United States data between 1948 and 1969 is illustrated in Figure 5.1.

In the next decades, many governments tried to use monetary policy to stimulate the economy. Suddenly, the Phillips curve, a stable relation over a century, disappeared. Inflation occurred with no gains in output or employment.