Protracted Frictional Unemployment as a Heavy Cost of Technical Progress

Neither neoclassical nor Keynesian economics displays much patience with the popular notion that technical progress of the labor-saving variety tends to swell the ranks of the unemployed. Those who believe that market forces tend automatically to bring the economy back, if not to "full employment," at least to a fairly stick "natural rate of unemployment" seem inclined to believe that this process will wipe out any joblessness created by technical change, presumably with some modest delay. The Keynesian approach suggests (subject to some recent concessions to the notion of the natural rate of unemployment) that the level of employment can be adjusted by macroeconomic policy and that this is capable of undoing whatever joblessness labor-saving innovation may engender. We will argue here that there is more substance to the public's fears that new productive techniques can threaten jobs than is acknowledged by these lines of analysis. We will suggest that when technical progress is a continuing process a speedup of change can have two profound employment effects. First, it can increase, perhaps materially, what used to be referred to as "frictional unemployment," thereby raising the natural rate of unemployment commensurately. Second, because of the sunk- cost attributes of the retraining of workers to enable them to use the constantly emerging novel techniques, speedup of technical change, rather than even-handedly leading to brief periods of unemployment to all of the workers affected, tends to single out three classes of workers--the ill-educated, the older former jobholders and women, particularly of childbearing age--either for declining relative wages or protracted and possibly for lifetime unemployment. There is, of course, a considerable body of writings on the social costs of economic growth. By and large it has emphasized the externalities generated by the growth process--crowding, damage to the environment, psychological tension, alienation and the like. It will be suggested here that the employment costs are arguably of at least comparable significance and that they must be taken into account more explicitly in any evaluation of a program dedicated to acceleration of economic growth