As indicated in the last section, the chief motivation for automation, as well as its chief effect, is to reduce the number of workers and save operating costs, while producing more goods. This is an illustration of technique at its best (or worst), for in this case the search for efficiency would clearly result in massive job displacement if taken to its fullest extent.
Whether the apparent material benefits are worth the disruption can be debated with good arguments on both sides. This situation however, does seem to illustrate the irresistibility of technique--even if one is capable of assessing the broader costs, automation will still take place because it produces more efficient results for the business. Also, the important ethical and social issues do not all lie at the start of the path, for the road is partly travelled already and the way back is cut off. Rather, they are found along the way, and relate to the appropriate responses that can be made to the process of automation. Only a few will be considered here.
It was remarked in the last section that relatively more of the future workers may be professionals, taking charge of their own education and training and contracting their services. Yet this route cannot be taken by everyone in the present-day work force. The typical assembly line or factory workers facing job-threatening automation will need considerable education and/or retraining to qualify for any new job, much less to take charge of their own destinies. Faced with a choice between unemployment and an arduous re-education, many will slide onto the welfare rolls, not as an active choice but as a passive one, for nothing in their background convinces them of the value of the harder path.
What is the ethical obligation of the other parties, including government and employers, to the large numbers of workers who are thus displaced? Surely the ethical imperative to assist others to be whole persons implies at least an offer to do the necessary retraining to allow re-employment. The employer may prefer, in consideration of the bottom-line profit, to simply terminate an unneeded worker. However, the months and years of employment have created a mutual bond and obligation (a social contract) that cannot exist between owner and machine but that always does between employer and employee. The employer who breaks this bond and discards the worker like a worn-out part creates bitterness and resentment that are certain to cost the broader society far more than job-retraining would have. The implicit social contract that the employer has with society as a whole binds both to act responsibly. Both therefore have an obligation to the person whose job has been automated to help make reasonable alternatives possible, or both will suffer the consequences of exaggerated class structure and broad social unrest (a pragmatic consideration). Government also has a responsibility to promote social stability, if for no other reason than (utilitarian) survival of the state itself.
The difficulty is that such responsibilities cannot easily be seen by employers, for they do not benefit the immediate bottom line, and they are hard to put into law. Some companies are too small or too unprofitable to afford such education. Yet unless retraining schemes are universal, a firm that does act responsibly in this way may become uncompetitive if others in their industry ignore those same responsibilities. Since life spans will probably increase in the future (see Chapter 7) and national economies will continue to change rapidly, the typical employee may need to retrain many times over a working lifetime. This projection also argues for a universal job retraining scheme, one in which employers, government, workers, and unions all participate.
A possible solution would be a comprehensive savings/insurance plan into which all parties pay--something similar to present-day pension and unemployment compensation schemes. If a job is automated, the employer could be required to increase payments to the plan for a period of time. On the other hand, after a certain number of years service, a worker ought to be able to take voluntary retraining at no additional cost, much as one might now take early retirement. Or, perhaps industry could learn from the sabbatical system used by academics to recharge their intellectual batteries every few years. After six years of service, a tenured university professor can normally apply for a one year leave at partial pay for the purpose of further study. Such plans have the advantage of recognizing the mutual obligation of all parties to retrain workers; they have the disadvantage of creating yet another payroll deduction and yet another administrative headache. Whatever it is called though, some such retraining insurance or educational pension plan may well be necessary in the light of events.
Such retraining plans do not provide the whole answer, however, for there would remain a core of workers who would be unwilling or unable to accept retraining. Since the newer jobs require more technical skills than the jobs being obsoleted, they also demand a more educated work force. For those who have held menial jobs because they could not do anything else, a technical education may not, in many cases, be a prospect.
No society could afford to have such people simply remain unemployed and collecting welfare the remainder of their working lives, for large numbers of jobless people have always been a destabilizing force in the past, and there is no reason to suppose that such a situation would not also lead to widespread rioting and destruction in the future. That is, the "haves" cannot for long wall themselves off from and ignore the "have-nots," for their own way of life is also at stake.
The number of unskilled labouring jobs in traditional sectors will continue to decline; the challenge is to find new jobs for those to whom it is not practical to give professional or technical educations. Since such jobs can only be created in the service sector, it is easy to predict great increases in, say, tourist-industry employment. This may be enough, but if it is not, there could be pressure to hire personal servants, estate caretakers, cooks, and maids--even to put human crews on farms or construction projects that could be done safer, faster, cheaper, and better with robots. There will also be some pressure to make intelligence enhancement devices and drugs (see Chapter 6), but it seems unlikely that these could soon be made universally available and thus they alone will not solve the problem.
The utopian ideal of some science fiction portrays every future citizen as idly rich, dabbling in professional activities while the robots do all the work. This vision is unrealistic even at current population levels, let alone at the higher ones that will soon prevail. If the expansion of service and information sector employment fails to absorb the displaced industrial work force, considerable creativity in job creation may be necessary to avoid massive social unrest. This problem could be severe even in the developed countries, testing the skill of the most democratic, honest, and caring of governments working with relatively prosperous citizens. In underdeveloped countries, the potential for disaster is great. It is not difficult to imagine the rise to power during a time of civil unrest of a tyrant who decides to rid his nation once for all of "undesirable" elements, slaughtering his own people in a dreary repetition of Nazi-like themes of racial or religious purity. It would also not be difficult to imagine wars between the "have" and the "have-not" nations or civil unrest in the prosperous ones. Change exposes fragility in social contracts, and it can cause more. The trick is to keep the fragile from breaking.
In the very long run, increased prosperity historically results in a significant decline in birth rates, but on a worldwide scale, no such remedy can be hoped for until several more decades have passed.
On a smaller scale, the methods chosen for introducing automation to a workplace, or for implementing any new system, can have a considerable effect on the morale and the jobs of the workers. The most common strategies are:
o Cold Turkey On the day assigned for the changeover to the new system, the old methods cease and the new replace them at once. Depending on the employer, this could involve either an extensive training program ahead of time, so that all the employees are ready to take up their new responsibilities, or a wholesale replacement of personnel. The latter approach is fraught with peril, not only for the workers displaced but also for the employer, who may find while awaiting the day of unemployment that employees do very little work or even engage in sabotage. Extensive retraining is also not without its difficulties, for even if the workers are able to take part in the actual planning of the new system, there are bound to be many errors in the period following the change. There may still be the problem of disgruntled workers, because automation is usually undertaken to achieve personnel efficiencies, and this implies that there will eventually be fewer workers. Moreover, as many organizations have discovered after such a change, few new products are either what they were advertised to be or are bug free, and most existing data sets have corruptions or anomalies that are revealed only in the changeover. In general, this is the worst possible way to introduce an automation or a new system.
o Phased Here, the change to a new system is made gradually, with some parts being operated in the old way while some are switched to the new. This method has the advantages of being less abrupt and disruptive and may be less error prone than the cold turkey changeover. However, there are inherent inefficiencies involved in partial automations or introductions, and the employer who does things this way must be prepared to wait until some time after the completion of the process to realize the expected gains. Indeed, since parts of the enterprise are operating under one system and parts under another, there may be some losses, for additional employees may be required during the changeover.
o Parallel This is the most costly of the three methods of implementing a new system. It involves running both the old and the new side-by-side for a period of time, comparing the results, and working out the bugs in the new through experience. In most manufacturing systems, the parallel method is not practical, and one of the first two methods must instead be chosen. However, in accounting systems, or student or employee record systems where great care must be taken to ensure the accuracy of the results at all times, it is unwise to place any trust in a new system until its results have been carefully compared to those of the old over a period of time. While such caution may be necessary in some cases, it may also imply the use of parallel staffs to operate the two systems. If the new staff is intended to replace the old, there will be great tensions; the old employees cannot be expected to work in the employer's interest during this time, for their energies will be put into finding new positions. If, on the other hand, the staff operating the new system is temporary and its other function is to train existing personnel in the new techniques, the atmosphere in the workplace may be better, though there may still be those whose jobs have been lost, and they are certain to resent the others. In general, parallel implementation, done properly, is the safest method.
Thus, the method chosen for automation can have a profound effect on the workplace. In the industrial age, the perception on the part of many employers was that the workplace belonged to them and they could do as they pleased with it, without regard to the effect on the workers. The perception on the part of many workers in the industrial age was that employers tended to exploit them, so they banded together in unions to create a power base of their own and to force changes in working conditions. The very idea of the information age, however, carries with it the assumption that workers will not only be fully informed about proposed changes but also that they as professionals will more often be in control of their own workplace. It will be they who will design and implement changes, for they will be the company. They will be less likely to be in opposition to the shareholders, as represented by the board of directors and officers, for they too will be shareholders and operators. In such circumstances, the problems associated with automation will not disappear, but the perception of them will, for change will not be something imposed from above; instead, it will be a result of the informed collaboration of professionals. Of course, conflict in the workplace will not be eliminated altogether by such changes; it will merely shift to other focal points.
Some of the considerations mentioned thus far sometimes lead to a questioning of certain basic assumptions most forecasters make--that automation is both desirable and inevitable. Perhaps it is neither, or only one, and not the other. Both benefits and problems are easily seen when considering the process in an abstract way. For the people directly involved, though, such dispassionate considerations may be impossible. Where automation is possible and does bring economic benefits, it will surely be done, whatever the other consequences (unless constrained by some higher authority). Perhaps it will be necessary to require that such consequences be examined in each case and specific provision made for displaced workers before proceeding. It may even be that most automations will proceed slowly enough that no serious large-scale problems develop, though that would not release individual employers from their obligations to workers.
It may also be that there are other solutions to the displacement problem that use yet-to-be-deployed technologies, and some of these possibilities will be examined in later chapters. It is clear that this second industrial revolution (automation) can no more proceed without social consequences than did the first, and that the choices involved in dealing with such problems are to a great extent ethical in nature.