A History of the Federal Reserve, Volume 2 by Allan H. Meltzer Page A

experienced a long period of sustained growth affecting more people in more countries than in any previous period. In the United States real per capita consumption more than doubled, more than an annual 2 percent compound average rate of increase. Although six recessions temporarily broke the economy’s growth, by the end of the 1970s inflation had become the major economic problem. At its peak rate of increase, consumer prices rose 12.5 percent in 1980 or 12.2 percent excluding one-time changes in food and energy prices (Council of Economic Advisers, 1989, 373). For the period as a whole, consumer prices doubled and redoubled, more than a 4 percent average annual rate of increase.
    Inflation was not at all uniform. It remained low until 1966, then rose with the financing of the Vietnam War and the Great Society. Inflation fell after 1984 and remained moderate in the late 1980s and beyond. Chart 1.1 shows these data. The Federal Reserve later chose to monitor the deflator for personal consumer expenditures excluding energy and food prices. In the short term this index differs from the consumer price index (CPI) mainly because weights on particular components differ in the two measures. Housing, medical care, and energy prices have been principal sources of short-term differences. Over a longer term, most broad-based price indexes move together. Many short-term differences result from large relative price changes, not from sustained inflation.
    Some economists attribute the Great Inflation to the end of price and wage controls or to the energy price increase. These changes affected the price level, temporarily raising the rate of price change. Persistent inflation shown in the chart, like all sustained inflation, resulted from excessive money growth.
    Productivity growth is a noisy series (chart 1.2). It is not possible to know promptly whether changes are transitory or likely to persist. Economists have not agreed on the reasons for the decline in productivity growth in the 1970s despite substantial research effort.
    Sustained changes in productivity growth are a main source of changes in expected output growth. Federal Reserve staff and many private sector economists use the difference between actual and expected output growth to forecast inflation. As Orphanides (2003a,b) showed, this was a main source of error in inflation forecasts in the 1970s. Meyer (2004) explained that it misled Federal Reserve economists who based inflation forecasts on the Phillips curve. Alan Greenspan did not rely on that model; his forecasts were more accurate in the 1990s.
    Chart 1.3 shows that output growth is highly variable. Some research shows that the series is closely approximated as a random walk—that is, a series dominated in the short term by its random component. This is a main reason that quarterly forecasts using econometric models lack accuracy.

    Currency depreciation follows domestic inflation unless foreigners inflate even more. Until August 1971, the United States kept the nominal exchange rate fixed at $35 per ounce of gold. Most other countries fixed their exchange rates in relation to the dollar and gold. To maintain the $35 dollar gold price, the Federal Reserve would have had to choose exchange rate stability over the requirements of domestic policy. The standard interpretation of the Employment Act of 1946 at the time gave most importance to maintaining full employment. Until the late 1970s full employment was considered a 4 percent unemployment rate. When the signals from the foreign exchange market and the unemployment rate diverged, the Federal Reserve followed the unemployment rate.
    The real exchange rate adjusts the nominal rate for differences in price levels at home and abroad. From 1951 through 1980 the real exchange rate declined persistently. In the early 1980s the United States ended high inflation. Chart 1.4 shows that from the mid-1970s on the real and nominal exchange rates moved together. Nominal exchange