Changes in the volume of unemployment are governed by three fundamental forces: the growth of the labor force, the increase in output per man hour, and the growth of total demand for goods and services. Changes in the average hours of work enter in exactly parallel fashion but have been quantitatively less significant. As productivity rises, less labor is required per dollar of national product, or more goods and services can be produced with the same number of man—goods. If output does not grow, employment will certainly fall if production increases more rapidly than productivity (less any decline in average hours worked, employment must rise. But the labor force grows, too. Unless gross national product (total final expenditure for goods and services corrected for price changes) rises more rapidly than the sum of productivity increase and labor force growth (again modified for any change in hours of work), the increase in employment will be inadequate to absorb the growth in the labor force. Inevitably the unemployment rate will increase. Only when total production expands faster than the rate of labor force growth plus the rate of productivity increase and minus the rate at which average annual hours fall does the unemployment fall. Increases in productivity were more important than growth of the labor force as sources of the wide gains in output experienced in the period from the end of the war to the mid-sixties. These increases in potential production simply were not matched by increases in demand adequate to maintain steady full employment. Except for the recession years of 1949, 1954, and 1958, the rate of economic growth exceeded the rate of productivity increase. However, in the late 1950s productivity and labor force were increasing more rapidly than usual, while the growth of output was slower than usual. This accounted for the change in employment rates. But if part of the national purpose is to reduce and contain unemployment, arithmetic is not enough. We must know which of the basic factors we can control and which we wish to control. Unemployment would have risen more slowly or fallen more rapidly if productivity had increased more slowly, or the labor force had increased more slowly, or the hours of work had fallen more steeply, or total output had grown more rapidly. These are not independent factors, however, and a change in any of them might have caused change in the others. A society can choose to reduce the growth of productivity, and it can probably find ways to frustrate its own creativity. However, while a reduction in the growth of productivity at the expense of potential output might result in higher employment in the short run, the long-run effect on the national interest would be disastrous. We must also give consideration to the fact that hidden beneath national averages is continuous movement into, out of, between, and within labor markets. For example, 15 years ago, the average number of persons in the labor force was 74 million, with about 70 million employed and 3.9 million unemployed. Yet 14 million experienced some term of unemployment in that year. Some were new entrants to the labor force others were laid off temporarily, the remainders were those who were permanently or indefinitely severed from their jobs. Thus, the average number unemployed during a year understates the actual volume of involuntary-displacement that occurs. High unemployment is not an inevitable result of the pace of technological change but the consequence of passive public policy. We can anticipate a moderate increase in the labor force accompanied by a slow and irregular decline in hours or work. It follows that the output of the economy—and the aggregate demand to buy it—must grow by more than 4 percent a year just to prevent the unemployment rate from rising, and by even more if the unemployment rate is to fail further. Yet our economy has seldom, if eve