provements are combined with increases in production volume. Effective labor compen- sation may rise or fall if productivity im- provements lead to shorter workweeks and/or new product prices, depending in large part on production volume and profit- ability. Finally, average wage levels will change with changes in the necessary mix of worker skills resulting from the implementa- tion of robotics and related technologies.
Definitions of unemployment, like those of productivity, require distinctions between short-term and persistent job loss, or be- tween true unemployment (job loss) and dis- placement (job shift).
For some time, most experts in the United States have argued that more jobs are cre- ated by new technology than are eliminated. However, if these jobs are in different in- dustries and/or require different skills, the effect on an individual who has been replaced by automation can be traumatic.
Production and servicing of robots and related technologies will create new jobs. The number of jobs created and the rate at which they appear will depend both on the growth rate of the robot industry and the degree to which robot manufacture and re- pair are, themselves, automated.
Additionally, the effects of modern micro- electronics will be to lower cost, improve per- formance, and widen the availability of automation technology substantially. Nega- tive impact on employment that, in the past, has been small enough to be insignificant or undetectable may be much larger in the future.
In order to assess the effects of automa- tion on future employment levels, a baseline must be established against which job loss or gain can be measured. This baseline could be a simple extrapolation of current trends. But it may also need to be adjusted to reflect two other effects:
Virtual employment, domestic jobs that
were not explicitly eliminated, but that
Ch. Ill—Social Issues
would have existed were robots not in- stalled. Virtual unemployment, domestic jobs that would have-been lost if the plant had not responded to domestic and in- ternational competition by automating.
As the case with productivity, it is dif- ficult to attribute employment effects to any single component of an entire range of im- provements in the manufacturing process, in this case robotics. Any examination of the effects of robots on jobs would need to con- sider, at least in part, a much broader con- text of automation technology.
There seemed to be two principal sets of questions concerning unemployment. These questions are different in their focus, in their implication for Federal policy, and in the data collection necessary to analyze them:
1. Will the United States experience a long-term rise in the real unemployment rate due to the introduction of robotics and other automation? If so, will these effects be differentially severe by geo- graphical location, social class, educa- tion level, race, sex, or other character- istics? What might be the employment penalty of not automating? 2 Will the use of robots create displace- ment effects over the next decade? In what ways will these effects be specific to particular industry classes, geograph- ical locations, or types of jobs? How will they effect labor/management negotia- tions?
Quality of working environment is another issue that was identified. If robots are employed principally for jobs that are un- pleasant or dangerous and if the new jobs created by robotics are better, the quality of worklife will improve. Productivity increases may also, in the longer term, result in a shorter, more flexibly scheduled workweek.
New forms of computer-based automation may in many cases relieve job boredom and resulting worker dissatisfaction that many management experts have been concerned