The Frontiers of Management. Peter F. Drucker

force, even though production might be 50 percent higher.

      If a company, an industry, or a country does not succeed in the next quarter century in sharply increasing manufacturing production, while sharply reducing the blue-collar work force, it cannot hope to remain competitive, or even to remain “developed.” It would decline fairly fast. Great Britain has been in industrial decline these last twenty-five years, largely because the number of blue-collar workers per unit of manufacturing production went down far more slowly than in all other noncommunist developed countries. Yet Britain has the highest unemployment rate among noncommunist developed countries: more than 13 percent.

      The British example indicates a new but critical economic equation: A country, an industry, or a company that puts the preservation of blue-collar manufacturing jobs ahead of being internationally competitive (and that implies steady shrinkage of such jobs) will soon have neither production nor steady jobs. The attempt to preserve industrial blue-collar jobs is actually a prescription for unemployment.

      On the national level, this is accepted only in Japan so far. Indeed, Japanese planners, whether those of the government or those of private business, start out with the assumption of a doubling of production within fifteen or twenty years based on a cut in blue-collar employment of 25 to 40 percent. And a good many large American companies such as IBM, General Electric, or the big automobile companies forecast parallel development. Implicit in this is also the paradoxical fact that a country will have the less general unemployment the faster it shrinks blue-collar employment in manufacturing.

      But this is not a conclusion that politicians, labor leaders, or indeed the general public can easily understand or accept.

      What will confuse the issue even more is that we are experiencing several separate and different shifts in the manufacturing economy.

      One is the acceleration of the substitution of knowledge and capital for manual labor. Where we spoke of mechanization a few decades ago, we now speak of robotization or automation. This is actually more a change in terminology than a change in reality. When Henry Ford introduced the assembly line in 1909, he cut the number of man-hours required to produce a motorcar by some 80 percent in two or three years: far more than anybody expects to happen as a result even of the most complete robotization. But there is no doubt that we are facing a new, sharp acceleration in the replacement of manual workers by machines, that is, by the products of knowledge.

      A second development—and in the long run it may be fully as important if not more important—is the shift from industries that are primarily labor-intensive to industries that, from the beginning, are primarily knowledge-intensive. The costs of the semiconductor microchip are about 70 percent knowledge and no more than 12 percent labor. Similarly, of the manufacturing costs of prescription drugs, “labor” represents no more than 10 or 15 percent, with knowledge—research, development, and clinical testing—representing almost 50 percent. By contrast, in the most fully robotized automobile plant labor would still account for 20 or 25 percent of the costs.

      Another, and highly confusing, development in manufacturing is the reversal of the dynamics of size. Since the early years of this century, the trend in all developed countries has been toward larger and ever larger manufacturing plants. The “economies of scale” greatly favored them. Perhaps equally important, what one might call the economies of management favored them. Up until recently, modern management seemed to be applicable only to fairly large units.

      This has been reversed with a vengeance the last fifteen to twenty years. The entire shrinkage in manufacturing jobs in the United States has been in large companies, beginning with the giants in steel and automobiles. Small and especially medium-size manufacturers have either held their own or actually added people. In respect to market standing, exports, and profitability too, smaller and especially middle-size businesses have done remarkably better than the big ones. The same reversal of the dynamics of size is occurring in the other developed countries as well, even in Japan, where bigger was always better and biggest meant best! The trend has reversed itself even in old industries. The most profitable automobile company these last years has not been one of the giants, but a medium-size manufacturer in Germany: BMW. The only profitable steel companies worldwide have been medium-size makers of specialty products, such as oil-drilling pipe, whether in the United States, in Sweden, or in Japan.

      In part, especially in the United States,* this is a result of a resurgence of entrepreneurship. But perhaps equally important, we have learned in the last thirty years how to manage the small and medium-size enterprise—to the point that the advantages of smaller size, for example, ease of communications and nearness to market and customer, increasingly outweigh what had been forbidding management limitations. Thus in the United States, but increasingly in the other leading manufacturing nations such as Japan and West Germany, the dynamism in the economy has shifted from the very big companies that dominated the world's industrial economy for thirty years after World War II to companies that, while much smaller, are still professionally managed and, largely, publicly financed.

      But also there are emerging two distinct kinds of “manufacturing industry”: one group that is materials-based, the industries that provided economic growth in the first three-quarters of this century; and another group that is information- and knowledge-based, pharmaceuticals, telecommunications, analytical instruments, information processing such as computers, and so on. And increasingly it is in the information-based manufacturing industries in which growth has come to center.

      These two groups differ in their economic characteristics and especially in respect to their position in the international economy. The products of materials-based industries have to be exported or imported as products. They appear in the balance of trade. The products of information-based industries can be exported or imported both as products and as services.

      An old example is the printed book. For one major scientific publishing company, “foreign earnings” account for two-thirds of total revenues. Yet the company exports few books, if any; books are heavy. It sells “rights.” Similarly, the most profitable computer “export sale” may actually show up in the statistics as an “import.” It is the fee some of the world's leading banks, some of the big multinationals, and some Japanese trading companies get for processing in their home offices data sent in electronically from their branches or their customers anywhere in the world.

      In all developed countries, knowledge workers have already become the center of gravity of the labor force, even in numbers. Even in manufacturing they will outnumber blue-collar workers within fewer than ten years. And then, exporting knowledge so that it produces license income, service fees, and royalties may actually create substantially more jobs than exporting goods.

      This then requires, as official Washington has apparently already realized, far greater emphasis in trade policy on “invisible trade” and on abolishing the barriers, mostly of the non-tariff kind, to the trade in services, such as information, finance and insurance, retailing, patents, and even health care. Indeed, within twenty years the income from invisible trade might easily be larger, for major developed countries, than the income from the export of goods. Traditionally, invisible trade has been treated as a stepchild, if it received any attention at all. Increasingly, it will become central.

      Another implication of the uncoupling of manufacturing production from manufacturing employment is, however, that the choice between an industrial policy that favors industrial production and one that favors industrial employment is going to be a singularly contentious political issue for the rest of this century. Historically these have always been considered two sides of the same coin. From now on, however, the two will increasingly pull in different directions and are indeed becoming alternatives, if not incompatible.

      “Benevolent neglect”—the policy of the Reagan administration these last few years—may be the best policy one can hope for, and the only one with a chance of success. It is not an accident, perhaps, that the United States has, next to Japan, by far the lowest unemployment rate of any industrially developed country. Still, there is surely need also for systematic efforts to retrain and to replace redundant blue-collar workers—something that