In the first half of 1995, I have counted over 50 nationally advertised meetings and symposia devoted to conjectures on the effects of the Internet. Newspapers and magazines now make the Internet headline news. Bookstores find Internet volumes their fastest-moving technical books. Adult education programs find that any course with the Internet in the title is a sure sellout, as a large contingent of information workers become apprehensive about their prospects for making a living off the information highway.
Even though much of this fuss is entertaining - if misinformed - computing professionals should not take it lightly. Whatever sociological or economic forces will spring forth from the digitization and so-called "demassification" of society, the readers of this journal are likely to be exposed to their effects sooner and more intensely than anyone else.
According to Malone's vision, independent consultants would do all systems work. Work would be outsourced to computer services corporations that would specialize in contracting for the individual contributors. Using the movie industry as a model, systems projects would be put together the same way movies are produced-using temporary teams of directors, camera people, and actors working on a rented set. The atomized systems professionals would be connected by multimedia Internet workstations. This would allow these individuals to offer their services in an electronic auction market. The allocation of resources would then be accomplished by competition and not by "management." I have labeled this model "electronically mediated mercenary outsourcing."
Just in case you think the MIT professors were not serious about visualizing work in the US as an adaptation of movie production, they explored the question of whether independent individual contractors could design and manage the production of automobiles. They found it plausible.
A student of management with a historical perspective will find the concepts of atomized individuals and deconstructed organizations strangely familiar. During the early stages of industrialization, jobs and enterprises were purposely designed so that workers could be retained only as long as they were ready, willing, and able to do what the owners wanted. When their usefulness waned, they were treated as machine parts and replaced by others who were more compliant, cost less, and became immediately productive without further training or capital investment.
Although the current visions are more sophisticated, they nevertheless fall into a pattern that has been historically proven to be destructive. However, there are important differences between the hire-and-fire industrial worker of yesteryear and the contemporary computer professional.
First and foremost, the computer budget of corporations is an overhead expense and consequently suspected of having no relationship to business productivity. In fact, survey after survey shows that computer professionals are identified with the rigidification of paper-shuffling bureaucracies and are therefore of questionable utility. Imposing marketplace disciplines on that lot by means of outsourcing is seen as a way to finally discipline an otherwise unmanageable group of employees. Creating industrial unemployment during the downside of an economic cycle never required such rationalizations.
Second, over 80 percent of the computer budget of corporations is for personnel expenses, most of which can be attributed not to doing measurable work but to learning what the organization wishes to do, attending meetings, and obtaining technical training on company time. In other words, systems professionals are paid to spend most of their time acquiring personal "knowledge capital." Most of that capital is company specific and will be lost when the professional leaves. But a large share of that capital is person specific and will represent an increase in the market value of the computer professional when he or she decides to take another job. Industrial employees never had such an advantage, since they rarely acquired any knowledge capital that would entitle them to increased wages.
In view of these differences, are the prophets of atomized individuals and deconstructed organizations realistic in reviving concepts of industrial organization for the information age? Or are the new proponents only reflecting their personal preferences as university professors and for-hire consultants?
Whether systems projects are like combat experience is debatable, although I find the similarities striking in view of intense global competition for markets. Based on a study of the Vietnam experience and how the U.S. Army has finally overcome it, I must reject the idea that enterprise integration, application development, and cumulative experience can be achieved by computer professionals who are seen merely as plug-in, plug-out software subroutines.
In fact, enterprises respond and operate much more as organic systems in which the performance of every element- regardless of how small-affects the performance of the entire body, just as the heart, lungs, stomach, brain, and glands are each indispensable for proper body function. In organic systems, all parts are essential for the health of the entire organism. When any one part fails to do its job properly, the system as a whole fails. Optimizing any part is not likely to improve the functioning of the entire system. Unless the new component is organically integrated with the other subsystems, it may actually inhibit overall performance.
The operative term for information systems is not optimization but balanced integration. What matters is not so much what each element does technically but how it fits with everything else. The unique characteristic of information systems is their pervasiveness. They should be seen not as isolated technical constructs but as carefully balanced functions that reflect the behavior of the entire enterprise.
Therefore I have grave doubts whether an overwhelming number of information professionals can find their most useful economic role as freelance infomercenaries who just get plugged in to do their bit of isolated work and then are dismissed. Information systems are the central element of how the modern enterprise governs itself, how it responds to external opportunities, and how it manages its resources. This requires a great deal of company-specific know-how, as well as a steady accumulation of knowledge.
Companies that think they can obtain economic value-added from information technologies exclusively in a bid/ask market, just like trading commodities, are likely to get just such commodity solutions. True, their labor costs will probably be lower, but they will totally miss the opportunity to capitalize on the accumulated knowledge of their people. It's just like the difference between home ownership and renting. After 20 years, the renters have only receipts for their monthly payments and no valuable assets.
The economic value-added of a model enterprise rests in the unique characteristics that distinguish it from its competitors. Standard solutions, therefore, will not give it a competitive edge, which is essential for not only earning current income, but also for generating sufficient funds for entering into new ventures.
The infomercenaries who think they can increase their earning power through for-hire short-term engagements will also be disappointed. If they are hired for their specific skills, as is likely to be the case, their wages will be made up of pay for work hours and an allowance for depreciation of their knowledge assets. The company hiring them will have no reason to finance their learning new skills. The infomercenaries will therefore have to look at their short-term wages as an exploitation of their accumulated know-how, which they will have to fund out of their own pocket to avoid technological obsolescence.
I have used the example of the banking sector-the most homogeneous one-to illustrate that there is absolutely no correlation between information technology budgets and profitability, regardless of what economic performance measure we choose. Information technology is related to input but not to output. The only plausible explanation for any differences is the manner in which any enterprise has taken a unique approach to combining its various elements of people, skill, motivation, and technology to deliver a result that differs from what competitors do with the identical resources.
If you view humans as machines, their component organs and chemical structure are very much alike and often surgically replaceable. Individuals-like organizations-are unique because of the way those parts are put together and how they manage their endowment, which increasingly is not physical but information based.
The prophets of the Internet, in conveying their vision of the future for professional workers, had better come up with scenarios that are supported by proven patterns of success. The year 2020 is not so far off that we do not already have clues as to what works and what doesn't. Information systems work is a means for an organization to achieve cooperation, integration, coherence, and uniqueness. I do not think that infomercenaries can deliver that.
 Garner, Rochelle. "Hello Cruel World." Computerworld (May 15, 1995), p. 83.
 This section was inspired by Russell L. Ackoff, The Democratic Corporation (New York: Oxford University Press, 1994).