KNOWLEDGE IN PRODUCTION [1]


       At a certain stage of their development, the material 
       productive forces of society come into conflict 
       with ... the property relations within which they 
       have been at work hitherto. From forms of development 
       of the forces of production these relations turn 
       into their fetters. Then begins an epoch of social 
       revolution. With the change of the economic foundation 
       the entire immense superstructure is more or less 
       rapidly transformed. -- Karl Marx [2]

       But if robots indeed are able to take the place of 
       human labor, critical questions arise. -- _San 
       Francisco Examiner_ [3]


INTRODUCTION

The designations "Information Age," "Second Industrial Revolution" 
and "Electronics Revolution" are all attempts at uniting under one 
banner the totality of recent developments in computers, digital 
telecommunications, robotics, bioengineering and materials 
science. Use of the word "revolution" recognizes the nature of 
these new technologies as qualitatively different from that which 
has gone before. 

Technology can de defined as the sum "of the means employed to 
provide objects necessary for human sustenance and comfort."[4] As 
such, "technology" comprises not only the machinery and tools 
required for production, but also workers' skills and the 
organization of production. In this paper, we hold that the 
technologies of this new era are distinguished from technologies 
of the industrial age by their high _knowledge_ content -- they 
can be characterized as knowledge-intensive. This distinction, 
which will be explored in more detail later, is warranted for two 
converging reasons. First, our widening understanding of Nature, 
especially in the life and material sciences, is yielding dramatic 
benefits in productivity, thereby reducing labor, machinery and 
raw material requirements in production. Second, the ability to 
_record_ workers' skills (another form of knowledge), and encode 
it using digital electronics into the instruments of production, 
and to _play back_ the knowledge in the absence of humans, reduces 
labor requirements in production. With the diminishing 
contributions of machinery and raw materials and labor, knowledge 
emerges as the dominating component in production.

We consider in this paper the implications for the organization of 
production based on knowledge-intensive technologies. As 
knowledge's role in production becomes dominant, it threatens the 
stability and viability of a system organized around the exchange 
of goods based on ability to pay.  Briefly put, the new 
technologies undermine current social relations. This paper pays 
particular attention to the effect on labor.

                                ***

For the purposes of this paper, we define _data_ as raw 
perceptions captured by some data-collection device. The Landsat 
satellite, for example, is capable of photographing the entire 
Earth's surface every two weeks, and has been operating for 20 
years. Ninety-five percent of the images have never been seen by 
human eyes.[5]

_Information_ is data with human labor applied to it. In its 
broadest sense, information includes experiences, perceptions, 
symbols, imagery, signals, and data that have been collected, 
organized, perhaps analyzed, and then expressed in some form. The 
key point here is that information is the product of human 
intellectual activity -- effort has been expended to put the data 
into a form capable of satisfying some need or want. Using the 
Landsat example,  the satellite photographs that have been 
examined and cataloged would qualify as "information." Information 
also includes transactions, customer lists, mail, news, research 
reports, and so forth.

_Knowledge_ is a further refinement of information. It too is a 
product of human labor. It is information that has been 
systematized and integrated, organized so that it is relevant to 
natural and social processes. Pursuing the Landsat example, an 
understanding of weather patterns, climate trends, mineral 
deposits, or land usage might be examples of knowledge derived 
from the cataloged and studied photographs. In this paper we focus 
on knowledge as a component of production.[6]


THE NATURE OF KNOWLEDGE IN PRODUCTION

Production cannot take place without knowledge. Some understanding 
of the production process is required so that production can take 
place. Just as knowledge enables production, more knowledge 
enhances it, i.e., it increases productivity. In one of its most 
obvious forms, additional knowledge, in the form of a worker's 
superior skill, enables him or her to accomplish a task more 
quickly, more easily, and with less waste. 

Knowledge in production takes many forms. In addition to being 
brought to the production process by the worker as _skills_, it 
might be contained in the _organization_ of production like the 
"assembly line" or the "work team," or in the _design_ of tools or 
machinery, or in the production _process_, or in the chemical 
_formulae_ or in the molecular _structure_ of a composite material 
or the DNA _sequence_ of a bioengineered protein, or in the 
software _algorithm_. Knowledge might be represented by the 
conservation _techniques_, or _methods_ of utilizing waste and 
by-products, or inventory management _theories_. 

Knowledge mobilizes the benefits of nature. Scientific and 
technological knowledge -- the universe as understood by society 
-- is Nature's bounty discovered.  It deepens or enhances or 
enlarges the environment in which production (and all other human 
activity) takes place. 

Knowledge has a _material_ basis. Knowledge cannot exist separate 
from some material "container" -- memory, books, computer disks, 
and so on. Knowledge, in and of itself, cannot create a house or a 
loaf of bread or a computer. Its usefulness only manifests itself 
through the production process. Knowledge can only generate useful 
things by having labor apply it during production. Knowledge can 
be disseminated, but be useless without the labor to apply it.

Knowledge has a _social_ origin -- it is the result of people 
interacting. Technology and invention are not the products of 
solitary inventors or scientists; rather, inventors and scientists 
build on the past accomplishments, experiences and discoveries of 
generations of scientists, engineers, authors, production workers, 
and so on.

Finally, we should note that knowledge has peculiar qualities 
which distinguish it from labor, machinery, raw materials, and 
other components of production. Two people can simultaneously use 
some bit of knowledge, it can be duplicated ad infinitum at almost 
no cost, it can circulate around the globe in seconds, it is not 
"consumed" or exhausted as it is used, and the more it is shared, 
the more it grows.[7] These qualities give knowledge a unique and 
subversive role in commodity production.

                                ***

Throughout history, humans labor, and as a result of that labor, 
society learns more about the Universe. The amount of knowledge 
increases. This knowledge shows up in technology. The development 
of the means of production is the accumulation of the experience 
of workers, refined by them into knowledge, and congealed in 
technology.

One of the general processes throughout the evolution of 
technology has been the transference of human skills and 
attributes to the machinery and techniques used in production. The 
need for humans as a source of physical power began to disappear 
with the domestication of animals and the harnessing of wind and 
water power. Later, the need for humans as a dexterous manipulator 
of materials began to disappear with the development of machinery 
that used gears, ratchets, springs, cams, etc. to replicate human 
motion. 

Most recently, the human functions of "operator" and "decision-
maker" have been usurped as the worker's knowledge is abstracted 
and programmed into numerical control ROMs, and the decision-
making is replaced by "artificial intelligence" programs.  This 
last stage has only been possible with the development of 
cybernetics, information theory, transistors, and a host of other 
technologies in the period around World War II. These technologies 
laid the basis for replicating functions of the human brain in 
inanimate electronic equivalents harnessed to machinery. With the 
invention in the late 1960's of low-cost programmable 
microprocessors and memory chips, knowledge could  be converted 
into sequences of computer code and processed by semiconductors at 
a speed of millions of simple instructions per second to direct 
machinery. In the past, the machine operator served as guide and 
overseer, steering and monitoring -- qualities difficult to 
reproduce using only mechanical means.  But the advent of new 
technologies allowed the machine operator's function to be 
abstracted and encoded, eliminating the need for an operator at 
all.  Thus began the assault on the last outpost occupied by 
humans in the production process.  

One key factor of knowledge-intensive production, then, is that 
the human aspect of production is "recorded", as it were,  
digitized, and _now capable of being "replayed" ad infinitum_. 
Knowledge becomes a "direct force of production."[8] What remains 
of the laboring process is the ever-shrinking pool of tasks like 
servicing or designing that are still beyond the dexterity or 
"programmability" of evolving technology; or tasks for which the 
cost of the automated machinery exceeds the labor required to 
carry it out.

The knowledge brought to mass manufacturing by workers (in the 
form of skills) or embodied in the machinery do not comprise the 
total required:  knowledge is also embedded in the organization of 
production itself.  The traditional assembly line was set-up once 
after which it ran without variation for month after month.  
Workers were consigned one task.  Change necessitated extensive 
revision and readjustment.  Current trends such as "lean 
production" are not hard-coded in this way but are intentionally 
malleable so that they can take advantage of new knowledge as it 
becomes available. This "new knowledge" appears as the outcome of 
ever-evolving experience with a particular manufacturing process; 
or as changes ordered by management; and -- a recent "innovation" 
-- the taking and rewarding of input from those who actually 
carry-out the work.  In the latter case, a "neotaylorism" emerges, 
where instead of the dictates coming down from above as they did 
in the past, workers in teams on the shop floor study ways to 
tailor themselves more tightly to the task.   

A production system with more knowledge applied to it will be more 
productive.  Differentials in labor, energy and material costs of 
course are important, but it is not these factors that are 
credited for the marketplace success, say, of Japanese 
manufacturers, the originators of 'lean' (i.e., knowledge-
intensive) production methods.  "A critical point was reached [in 
the development of production technologies] when the American 
automobile industry finally acknowledged that Japanese firms had 
established a real production advantage based on the more 
effective use of capital -- that is, in the effective organization 
of labor and equipment -- not on low-cost labor."[9]

                                ***

In knowledge-intensive production, significant amounts of 
scientific research (mental labor) are carried out outside of and 
prior to actual production. The product of research is then 
brought to production in the forms of designs, new materials, 
techniques, algorithms, biotechnologies, etc.  The products of 
this scientific effort employ a deeper understanding of natural 
processes, down to the molecular and atomic level.  Some new 
materials "remember" qualities when exposed to light or 
temperature changes; others are super-conductive; still others 
facilitate data transmission at the speed of light.  Toffler 
describes this leap in productive forces: "Second Wave industries 
used brute force technologies -- they punched, hammered, rolled, 
beat, chipped and chopped, drilled and battered raw materials into 
the shapes we needed or wanted ... The Third Wave industries 
operate at an altogether deeper level. Instead of banging 
something into shape, we reach back into the material itself and 
reprogram if to assume the shape we desire."[10]  As a result, 
"new materials and biotechnologies, along with information 
technologies, undercut the value of many existing sources of 
natural resources, since they allow the replacement of one 
material with another, or permit the more efficient use of already 
available objects."[11]  Intellectual work beforehand substitutes 
for material (and labor, as described above) in production.  This 
has always been the case, but the rate of substitution has 
accelerated, and its application is facilitated by new 
technologies (e.g., telecommunications technologies disperse them, 
and flexible manufacturing methods enable their ready 
implementation).

Developments in science and technology reinforce the fact that we 
have made a radical break with our industrial past. Electronic 
technology and cybernetics enable instructions and information to 
be coded in digital electrical pulses, instead of gears, ratchets 
and springs. The result is components  (e.g. digital communication 
switches) that do not suffer from friction or fatigue, and operate 
at speeds and load levels orders of magnitude greater than their 
mechanical counterparts.  A constantly repeating sequence of 
instructions does not wear out a processor chip the way repeated 
actions eventually wear out a piston or camshaft.[12]  Beyond 
solid-state electronics, the developing field of "molecular 
electronics" utilizes the properties of proteins and bacteria in 
production.[13] The possibility of bacteria tirelessly creating 
polymers in a laboratory vat (now a possibility), without the cost 
of locating, drilling, pumping, transporting and processing oil to 
achieve the same result suggests a level of productivity that is 
qualitatively different from previous forms of productive forces.   

Or consider the particular case of the computer software industry. 
Software, the instructions that direct machinery, is knowledge 
encoded.  It is distilled experience and learning listed out in 
instructions for hardware.  Forty years ago, there was no software 
industry. In 1990, the global market for packaged software was 
estimated at $43 billion, and is expected to reach $100 billion by 
1995.[14]

But where does software stop and hardware begin? The distinction 
becomes fuzzy when machinery incorporates chips with instructions 
"burned into" them. Chip masks for semiconductor production have 
been legally defined as a form of writing, and once a mask of 
acceptable quality is developed, becomes something like a printing 
plate for reproducing chips (and money). Even the metals and 
composite ceramics used in production are now spoken of as being 
"smart", as in, "[t]he smarter the material, the less it 
weighs."[15]


                                ***

In examining the nature of knowledge in production, we might look 
for quantitative measurements to justify calling the new 
technologies "knowledge-intensive." Knowledge (admittedly a soft 
term) can take the form of "already known" knowledge, in the form 
of existing technology, training, education, copies of software, 
etc., or of "newly known" knowledge, the results of research and 
development. 

The extensive training required to master new technologies, and 
the need for ongoing education to keep current with a given field 
indicates the volume of "already known" knowledge required for 
contemporary production. Or one might use the number of 
instructions built into a machine in a state-of-the-art factory, 
or years of education per worker to measure "knowledge-intensity."

One measurement of "newly known" knowledge might be the relative 
portion of _design_ effort that goes into a product, versus the 
actual _production_ effort. "[I]nformation, design, research and 
development and software represent a growing proportion of the 
value of most products," note the authors of _Beyond the Casino 
Economy_, an analysis of Great Britain's economy. "As the 
importance of research and development rises relative to that of 
direct production, the purpose of labour is increasingly the 
production of knowledge, in the form of designs or production 
processes."[16]  The recording of knowledge in the form of 
software makes up ever greater proportions of pivotal tools and 
industries.  Software becomes a larger and larger cost of 
increasingly complex production systems: "Software now accounts 
for about 80 percent of the development expense for new systems," 
according to _Computer Design_ magazine in 1987.[17] Even in 
manufacturing, software plays a central role. "Retooling", with 
the new "flexible manufacturing systems", simply means changing 
the software that guides the machines. The assembly-line 
(hardware) remains unchanged.  The robots, hardly pausing, begin 
exercising different actions in obeyance of the newly-loaded 
programs.

Scott Lash has suggested that the number of models per design 
serves as a crude indicator of the relative amount of knowledge 
that goes into a good. I.e., a model of a car of which only 10,000 
"copies" are made has a higher design or knowledge content than a 
model of which 20,000 are made.[18] The trend in manufacturing has 
been towards smaller production runs, with more frequent model 
changes, suggesting therefore a higher design/knowledge content.  

Finally, we might look at employment statistics as an indicator of 
the changing character of the productive forces. By 2000, one 
study estimates that two-thirds of the employed workforce in the 
U.S. will be working in education/knowledge/information-related 
jobs, while manufacturing, commerce and industry will account for 
only 22%, and agriculture for 2%.  In 1920, 9% worked in 
information/knowledge/education jobs; 53% in manufacturing; and 
28% in agriculture.[19] But the lines between these categories are 
blurring as the knowledge component in each sector increases. In a 
report on high-tech tractors, a modern farmer was quoted, "Farming 
you used to do with your back; now you use your mind."[20]


KNOWLEDGE IN THE PRODUCTION OF COMMODITIES

A common problem in discussing "knowledge" as a factor in 
production is determining its "value," and what value it adds to 
goods during production. Toffler, for example, says that 
"knowledge adds value."[21] But what _is_ "value"? An economics 
textbook defines "value added" as simply "the revenue from selling 
a product minus the amounts paid for goods and services purchased 
from other firms."[22] This definition is unsatisfactory. Is 
"value" only realized through the "selling" and the "purchasing" 
-- in the realm of circulation? What about the production process? 
Is value really only tied to the vagaries of fluctuating supply 
and demand?  What if the "goods and services" can't be sold, say, 
because potential users do not have the money to purchase the 
product? Does the product therefore have less (or no) value?

Recognizing the central role of commodities in capitalism, Marx 
began _Capital_ with an extended analysis of the question of the 
"value" of commodities. He identified two different kinds of 
"value" in commodities. In order to be exchanged, a commodity must 
fulfill some need or want for another human being. Marx called 
this subjective and qualitative aspect of a commodity its _use 
value_. At the same time, in order to exchange goods of different 
use values, Marx argued that there needs to be some common basis 
of assessing a value of the commodities, some quantitative, 
measurable aspect. Marx identifies "socially necessary labor" as 
that "thing" common to all commodities. It represents the amount 
of abstract human labor added during production, and the "dead" 
labor embodied in the raw materials and machinery used up during 
production. Marx called this aspect of commodities _exchange 
value_. The purpose of production, the reason that humans come 
together and engage in economic activity, Marx argued, is to 
create use values, to satisfy needs and wants. The process of 
production, however, is the expenditure of past and present human 
labor, measured as exchange value. The exchange value of 
_knowledge_, then, is the "socially necessary labor" that goes 
into the research, the analysis, and the expression required to 
develop it.[23]

Marx  defines "socially necessary labor" as "that required to 
produce an article under the normal conditions of production, and 
with the average degree of skill and intensity prevalent at the 
time."[24] The concept of "socially necessary labor" that defines 
the exchange value of a commodity recognizes an "average" 
technology stage or platform upon which production takes place. 
The "socially necessary labor" then, implies also a certain common 
level of knowledge about production processes. The use of 
computerized typesetting in newspaper production, robotics in 
automobile manufacture, or crop rotation in agriculture are 
examples of a technology platform.  Some producers may be ahead of 
the average, because of some special knowledge or technique, and 
some may be behind the average, because they are unaware of a 
technique, or have not invested in state-of-the-art technology. A 
commodity made by a worker employed by the "behind the average" 
company with outdated technology or using outdated techniques does 
not have more value because the worker took longer to make it. 
_Nor does the commodity have less value if an especially 
productive worker, using state-of-the-art equipment with the 
latest techniques takes less time to make it_.

In the latter example, a capitalist enterprise can realize extra 
profit from use of some particular knowledge as long as the 
knowledge enables him to produce commodities whose value is less 
than the "average" value of that commodity from all producers, 
both slow and fast, both backward and advanced. The advanced 
producer's commodities contain less than the socially necessary 
labor -- the enterprise ahead of the innovation wave is producing 
commodities more cheaply than its competitors, but selling them at 
the same price on the market. Thus, certain kinds of knowledge 
become sought-after resources; and competition drives forward 
technological development, although in a haphazard and socially 
hazardous way, because maximum profitability is the overriding 
goal.

Once knowledge becomes the new social average (that is, it becomes 
widely disseminated so now everyone is using the new technique), 
its ability to enable the innovator to accumulate extra profit is 
lost. To _maximize_ profit from knowledge, then, the capitalist 
must enjoy the _exclusive_ use of it.

In order to preserve the value of knowledge for the originator, 
knowledge used in production must be contained, and prevented from 
becoming the social average. The innovator tries to keep new 
techniques that give the firm an advantage hidden from 
competitors. At the same time, however, competing capitalists want 
to get hold of the newest technology to effectively compete. The 
patent and copyright system was developed, and continues to 
develop through laws and the courts, to attempt to resolve these 
two contradictory demands by competing capitalists -- _protection 
of profit_ (protecting the producer of the knowledge or 
technology) vs. _access to profits_ (access by competitors who 
want the knowledge or technology).[25] Copyrights and patents are 
the legal mechanisms for maintaining exclusive rights to a 
particular technique. They are treated as assets on company 
balance sheets, and represent sources of revenue, like mineral 
deposits or trade routes or right-of-ways.

The economics of "knowledge production" are such that the initial 
version requires a substantial investment (a high fixed cost), but 
subsequent copies have a relatively low reproduction cost.[26]  
Thus, the exclusive, original copy of the knowledge has high 
exchange value. But just as machinery loses value as cheaper 
versions come into use, copies of knowledge, because of the 
relatively low cost of duplicating knowledge (hence cheaper 
versions of the original), quickly depreciate the exchange value 
of the original knowledge.[27] For subsequent users, the 
knowledge, once it's become the social average (i.e., widely known 
or distributed) continues to add to the mass of use values, but 
transfers little or no exchange value to commodities in the course 
of production. Each copy (book, computer disk, tape, etc.) of 
"knowledge" consumes almost no material relative to its 
development cost, so has little exchange value to transfer to the 
final product.  Compare this with, say, a machine cutting tool.  
Each "copy" of the cutting tool consumes additional steel, energy, 
labor, and so forth, so it may have a substantial exchange value 
to transfer to the final product.

A century and a half ago, Marx noted that "[a]ll means of 
production supplied by Nature without human assistance such as 
land, water, metals in situ, and timber in virgin forests" fall 
into a category of things which transfer use value, without 
transferring exchange value.[28] Elsewhere, Marx referred to the 
"gratuitous" work of machines, the result of the machinery 
mobilizing natural forces.[29] He also recognized that "the 
productive forces resulting from cooperation and division of labor 
cost capital nothing. They are natural forces of social 
production. So also physical forces, like steam, water, etc. when 
appropriated to productive processes cost nothing."[30] 
"Cooperation" and "division of labor" -- learned ideas of how to 
organize production -- are examples of knowledge. Once discovered, 
knowledge costs nothing (i.e., transfers little or no exchange 
value), but enhances productivity, and thus adds to the mass of 
use values. This is the character of contemporary productive 
forces. So when Toffler says that "knowledge adds value," he is 
correct in the sense that it adds to the mass of _use_ values. But 
in another sense he is wrong, because knowledge reduces the 
_exchange_ value of commodities.

Adding machinery to production increases the constant portion of 
capital. It is development based on expansion of requirements -- 
more raw materials, more fixed capital. Knowledge, on the other 
hand, _reduces_ the constant portion of capital and production 
requirements, while at the same time _expanding_ output. The cost 
of computing power, for example, has plummeted because of new 
materials and new designs.[31] Miniaturization, computerized 
controls, conservation techniques and new composite "smart" 
materials reduce raw material and energy requirements in 
manufacturing and agriculture. Computerized inventory control and 
digital telecommunications reduce inventory requirements and speed 
the turnover of capital. Some economists assign a majority, and in 
some countries, over 75%, of the postwar economic growth in the 
West to improved productivity via technology, as opposed to growth 
resulting from increased inputs like more labor, raw materials and 
machinery.[32] Knowledge, as a special form of information, now 
dominates production itself, and overwhelms the contributions from 
traditional inputs to the final product.[33]


THE IMPACT ON LABOR

The twofold process we have been describing -- knowledge-intensive 
production is able to "record" and "play back" human effort many 
times in the absence, for all practical purposes, of human beings; 
and at the same time, knowledge mobilizes the "in situ" benefits 
of Nature, extracting use values possessing little or no exchange 
values -- points to the elimination of labor from production. 
Exchange value is the measure of human labor power consumed in 
production. A drop in exchange value represents a drop in human 
labor requirements. 

Concretely, new technologies affect the labor market in different 
ways. Forester identifies four "causes of concern" regarding the 
impact of high technology on jobs. These include (1) traditional 
manufacturing jobs (e.g., in automobile and steel production) are 
disappearing, and will never come back; (2) new manufacturing 
industries will not create many new jobs, because of automation; 
(3) there are doubts about the capacity of the service sector to 
create any more new jobs; and (4) the high tech sector itself, 
even though it might grow, will create only a modest number of 
jobs.[34]

Under capitalism, one of the primary reasons for introducing new 
technology is to reduce costs (towards maximizing profit), 
including labor costs. The greater the savings, the greater the 
incentive to innovate. Robotics and numerical control technology 
enables firms to eliminate high-paid production jobs. In a recent 
interview, Heidi and Alvin Toffler describe a new factory in 
Israel. It "is a cutting tool factory that doesn't have a single 
worker in it. Even machines get their parts from a robot that goes 
from machine to machine and resupplies them." What work is done is 
moving information around either as "technical specialists" or 
helping the products circulate, "working in bookkeeping or in 
sales."[35]  The Next computer factory in Fremont, California 
"requires only five manual-assembly workers and fewer than a 
hundred other workers, mostly engineers, for a line capable 
producing $1 billion of computers a year."[36]

Better data access and analysis as a result of computers, improved 
telecommunications, and networking enable firms to also eliminate 
middle-management positions. Digital telecommunications, improved 
transportation technology, and modern manufacturing methods enable 
the globalization of production and the labor market. This makes 
it economically feasible to transfer work (that can be moved, like 
manufacturing, as opposed to work that cannot be moved, like 
personal services) to cheaper labor markets; and at the same time 
squeezes domestic wages. Recent figures from the U.S. Bureau of 
Labor Statistics indicate that the same number of people will be 
employed in manufacturing in 2000 as in 1970. This means that the 
number of people employed in manufacturing, as a percent of total 
employment, will shrink from 24% in 1970 to 14% at the turn of the 
century. From 1970 to 1988, however, manufacturing output has 
remained a steady 20% or so of the GNP, while the output, in 
constant dollars, has grown by more than 50%. Other figures 
indicate that, over the last ten years alone, 1 million 
manufacturing jobs have disappeared in the U.S.  In 1982 dollars, 
the average weekly earnings (including overtime) for private 
industry production and non-supervisory workers was $298 in 1970; 
in 1989 it was $264, an 11% drop.[37] Fewer manufacturing workers 
are producing more, and in general, making less.

At the same time a relatively small, well-paid, knowledge-rich 
(highly skilled) section are still eagerly sought after by firms. 
"Not all parts of the labor market are shrinking. Engineers and 
technicians are still in demand, depending on their specialty, 
even at companies that otherwise are paring workers."[38] The 
working class is splitting into a well-paid, knowledge-rich 
(highly skilled) section that works in capital-intensive design 
and production work, and a larger, relatively low-skilled (or no-
skilled) section consigned to work that is too expensive to 
automate, with an ever-widening gulf growing in between.[39] The 
polarization of income has been dramatic. A recent _Business Week_ 
article points out that "it's only those in the top 20% who show a 
respectable gain in real incomes over the 15-year span." The 
bottom 60% have seen their incomes drop, while the richest 5% have 
seen their income grow 60%, while the richest 1% have seen rates 
of growth twice that.[40]

Whole sections of the U.S. population are being cast out of the 
sphere of production. The cast-outs are neither consumers nor 
producers; they aren't even needed as a "reserve army of the 
unemployed."  Drugs, disease, illiteracy, homelessness, or prison 
are their lot. "High-wage slavery is being replaced by low-wage 
slavery.  Low-wage slavery is being replaced with no-wage slavery 
-- people who work without wages, but who are 'paid' in survival 
coupons."[41] The members of this latter group are marginally 
maintained by society through shrinking welfare payments and food 
stamps, and forced to earn their meager keep through modern 
versions of slavery like workfare or prison labor.[42]  The "cost 
of production" of marginalized workers exceeds their usefulness as 
laborers -- in the logic of capitalism, they are people with no 
value.[43]

The "well-paid, knowledge-rich" section of the working class is by 
no means immune from the pressures on wages and "redundancy" 
originating in the information economy. Behind the drive for 
object-oriented programming (OOP), for example, is the need to cut 
costs and raise productivity by bringing software production 
techniques at least to the level of the interchangeable part -- 
something achieved in industrial production 150 years ago. A 
_Business Week_ editorial argued, "There's already evidence that 
object-oriented programming can help corporate computing 
departments reduce the outlandish amounts of money and time spent 
on creating their own programs. This could spell substantial 
savings, since corporations now spend most of their information-
technology budgets on software -- about 60% more than they spend 
on hardware, according to market researchers."[44] The next 
technological step beyond OOP is computer-aided software 
engineering (CASE), which could bring software production up to 
the electronic age, by having computers themselves write the 
software.[45]

The current recession has pummeled the electronics industry, as 
much as any other. IBM has announced that an additional 20,000 
jobs will be cut in 1992, on top of the 20,000 cut in 1991. IBM 
will have eliminated 75,000 jobs -- almost 20% of its workforce -- 
since 1986. Some 90,000 jobs were lost nationally in the 
electronics industry during the year ending in September, 
1991.[46] The mini- and mainframe computer companies like Digital, 
Bull, Wang, Burroughs, Tandem, Amdahl and IBM have been victims of 
the rapid "downsizing" (replacing large, old computer systems with 
small, cheap, more powerful systems) in hardware and software 
driven by new technology and the recession.[47] Under the 
euphemism of "restructuring", workers in these technology firms 
are being cut loose, ironically due, in part, to even more 
powerful computer technology.

There is also no reason why data entry, computer programming or 
data analysis cannot be done in low-wage areas like India, 
Ireland, or Eastern Europe, with the product of the labor, 
computer code or data, transmitted instantaneously electronically 
to customers on the other side of the world.[48] A recent _Wall 
Street Journal_ article described how data processing and other 
"back-office work" is being moved offshore to cheaper labor 
markets. It's not only low-skilled data entry work that's moving. 

       Wright Investor Services has 85 employees in its 
       Shannon [Ireland] office. Most of them are young 
       financial analysts earning less than the equivalent 
       of $20,000 a year organizing financial information 
       from companies around the world for Wright's 
       databases. That is far preferable to hiring 
       American business school graduates at $45,000... 
       It is precisely these kinds of higher-level jobs 
       -- financial analysts and technicians -- that 
       the Irish government is trying to attract. And 
       much the same can be said for Jamaica, Singapore 
       and, for that matter, many U.S. communities. So 
       the future may bring intensified worldwide 
       competition for these high-skill computer-based 
       tasks.[49]

The article goes on to describe how software developer Quarterdeck 
employs 20 workers in Ireland to field nearly a thousand technical 
support calls a week. During the day, calls are handled by the 
U.S. staff, "but after hours, the head office [in California] 
throws a switch and the calls are routed automatically to 
Ireland." Intercontinental Software in Palo Alto, California, 
founded by a Bulgarian emigre, brokers well-trained but relatively 
inexpensive East European programmers for American firms seeking 
to lower software development costs.[50]


CONCLUSION

Knowledge costs almost nothing to duplicate, especially if it 
appears in digital form. As a greater percentage of goods become 
knowledge, the nature of production as resource-exhaustive, labor-
consuming, and scarcity-bound becomes obsolete. The new productive 
forces are resource-conservative yet generate an abundance. 
"Ownership" becomes an irrelevant concept if many people can 
possess the same thing simultaneously. Property rights as we have 
known them simply get in the way, and hold back development. The 
holding back takes many forms: incompatible standards that 
needlessly complicate learning new skills and sharing information; 
unnecessary and wasteful duplication of research and development; 
expensive lawsuits, ultimately paid for by the consumer, over 
ownership of interfaces; increased surveillance to catch 
"information pirates"; decreased access to public information as 
databases are privatized and information is commodified; skewed 
priorities as profitability and not social need are the 
determining factor in research, development and distribution of 
knowledge; and even the criminalization of knowledge itself as it 
is classified as weaponry, lest it get into the wrong minds.[51] 
Society is harmed, and social development is held back.

With fewer jobs and lower wages as a result of the new knowledge-
intensive forces of production, the circulation of commodities in 
exchange for wages becomes impossible. Wages are simply not high 
enough nor extensive enough to absorb the productivity of the 
economy.  Private property laws separate the destitute worker from 
the means of survival. So apartments sit empty, while homeless 
people sleep in doorways or in prison-like shelters.[52] Food sits 
in warehouses or is destroyed, while children suffer from 
malnutrition. Illiteracy rates climb while teachers are laid off.  
Meanwhile, the capitalist scrambles to protect his position by 
locking up knowledge,  by looking for new areas to commodify and 
convert into sources of profit, and by further revolutionizing 
production.  More workers are laid off, or jobs eliminated through 
"early retirement;" this only exacerbates the crisis.

A. Sivanandan describes the revolution in technology as 
"emancipating" capital from labor. 

       [T]he more Labour tries to hold Capital in 
       thrall by withholding its labour, the more Capital 
       moves towards its emancipation through yet more 
       information technology, yet more labour-less 
       productive regimes, yet more recourse to the 
       captive labour force in the periphery. The relations 
       of production, that is, have changed with the 
       changes in the level of the productive forces: 
       information (in the sense of data fed to computers, 
       robots, etc.) increasingly replaces labour as a 
       factor of production; Capital no longer needs 
       living labour as before, not in the same numbers, 
       in the same place, at the same time; Labour can 
       no longer organise on that basis, it has lost 
       its economic clout and, with it, whatever political 
       clout it had, whatever determinancy it could 
       exercise in the political realm... And this is 
       what moves the battle from the economic to the 
       political...[53]

The problem becomes not how to produce wealth, but how to 
distribute it. As such, the struggle is not around wages, or job 
security per se -- economic struggles -- but around property 
relations and social relations, around the social contract and 
social convention of ownership,  around social control and 
survival  -- political issues.

A syndicated article by Robert Lewis appeared in November 1990, in 
the _San Francisco Examiner_  (ironically, in the employment want 
ad section) with the headline "Will the age of the robots produce 
a workless society?" 

       Imagine a society where material needs are provided 
       by "smart" machines, where people manage to 
       break the link that equates self-worth with a 
       job and are able to live comfortably from the 
       fruits of robot labor...

       Computer scientists Hans Moravec of Carnegie 
       Mellon University and Kalman A. Toth, founder of 
       the Silico-Magnetic Intelligence Corp., predict 
       robots will be commonplace in 10 years. In 50 
       years they say, robots will have replaced most 
       if not all human labor...

       Experts say the widespread entry of robots into 
       the workplace could raise the living standards 
       unlike any invention during the industrial 
       revolution. But if robots indeed are able to 
       take the place of human labor, critical questions 
       arise.

       First, how should the wealth produced by enterprises 
       operated with robot labor be distributed to 
       those who don't work or who work part time? Toth 
       says he envisions that non-workers would receive 
       "citizen pay" on a basis that would have to be 
       worked out...


Between people's needs, and the immense productivity of the 
knowledge economy, stands a system of property relations. These 
relations are historical -- "private property," as a social 
convention, developed, not without much struggle during the 
beginning of the capitalist period in the 16th and 17th 
centuries.[54] Such a system of property relations was required 
for private ownership of means of production, and the protection 
of newly acquired wealth from both the feudal powers and the 
emerging "property-less" classes. 

There is nothing "natural" about property rights, nor are they 
universally recognized.[55] Rather, they are conventions struggled 
over, formally and informally, by various social forces. Different 
sections of society respond to these developments in different 
ways. Among the most destitute section of society, it takes the 
form of struggling to open empty HUD houses to the homeless, to 
distribute food in government warehouses to the hungry, or to 
provide livable welfare grants by raising taxes of the wealthy.

Marx and many other writers have pointed out that social relations 
eventually must correspond to the level of productive forces. We 
now live in a time when productive forces have raced far ahead of 
social relations. The knowledge-intensive productive forces are 
straining against the chains of private property relations. The 
qualities of knowledge, to be fully maximized, require a system 
based on cooperation and sharing, because cooperation and sharing 
generates more information and social wealth. Such a system would 
emphasize education, because education builds the infrastructure 
for expanding social wealth. Such a system would require the 
distribution of goods on the basis of need, because the cost of 
production eliminates scarcity and wages. This, of course, is a 
radically different system. Then again, the technology we use to 
produce goods now is radically different.

--------------------
FOOTNOTES

1. A version of this paper originally appeared in _Directions and 
Implications of Advanced Computing 1992 (DIAC-92)  Symposium 
Proceedings_, Computer Professionals for Social Responsibility, 
Palo Alto, CA.

2. Karl Marx, "Preface to _A Contribution to the Critique of 
Political Economy_." In _Selected Works_. p. 182.

3. R. Lewis. "Will the age of robots produce a workless society?" 
_San Francisco Examiner_. November, 1990.

4. _Webster's New Collegiate Dictionary_ (1975).

5. A.Gore, "Infrastructure for the Global Village." _Scientific 
American_. September, 1991. p. 151.

6. There are certainly other important aspects of information in 
today's economy. Large quantities of information are required to 
circulate goods in a global market; and information itself has 
become a commodity  (D. Schiller, 1988). The reproduction of labor 
power requires transmitting the knowledge required to continue 
production, through the educational system and publishing. 
Entertaining and propagandizing to preserve the social order is 
handled by and large through "information industries," including 
the news media and entertainment industry. Methods of coercive 
social control (via the police and military) keep pace with 
technology as well.

7. Cleveland; also Toffler, _Previews and Premises_, p. 25.

8. "Nature builds no machines, no locomotives, railways, electric 
telegraphs, self acting mules etc. These are products of human 
industry; natural material transformed into organs of ... human 
participation in nature. They are _organs of the human brain, 
created by the human hand_; the power of knowledge objectified. 
The development of fixed capital indicates to what degree general 
social knowledge has become a _direct force of production_..." 
Karl Marx, _Grundrisse_, p. 706. [emphasis in original]

9. S. Cohen and J. Sysman, p. 118.  See also footnote below, which 
indicates that manufacturing output in the U.S. continues to climb 
with fewer workers.

10. Toffler, _Previews and Premises_. p. 20. Toffler uses "waves" 
to describe in broad strokes these stages: the First Wave 
corresponds to the era of primarily agricultural-based manual 
labor; the Second Wave corresponds to mechanized industrial 
production; the Third Wave corresponds to the knowledge-intensive 
production based in electronics, biotechnology, and new materials 
(see _The Third Wave_).

11. Goldhaber,  p. 61.

12. League for Programming Freedom,  "Against Software Patents."

13. "A protein isolated from bacteria found in salt marshes is 
proving to be a promising device for data storage in a molecule... 
A group of researchers from Syracuse University reported they 
optically stored and retrieved data in three dimensions in a tiny 
block made of molecules of bacteriorhodapsin... " The article goes 
on to say that six 1-cm. cubes of the molecules can store the 
entire Library of Congress. Amal Kumar Naj, "Bacteria Protein May 
Help to Miniaturize Computers." _Wall Street Journal_. September 
4, 1991. p. B-4.

14. _1991 U.S. Industrial Outlook_. U.S. Department of Commerce, 
January, 1991. p.28-15. These figures are for the "packaged 
software market", and do not include computer programming services 
or in-house computer programming.

15. Dieter Altehpohl, an executive with Alusuisse, quoted in 
Cleveland.

16. Costello, Michie, and Milne, 1989. A recent _Business Week 
_pointed out "automation has whittled hands-on labor to 15% of 
manufacturing costs, and in high-tech industries, it's closer to 
5%." (Kelly, 1991)
 
17. Quoted in Hayes, p. 87.

18. S. Lash, 1991.

19. _New Jersey Bell Journal_, 1984, quoted in the Cleveland.

20. S. Rosenfeld, "'Smart tractors' transform farming," _San 
Francisco Examiner_, June 10, 1991.

21. _Powershift_, p. 82.

22. Baumol and Blinder, 1979. p. 322. 

23. Workers who process information can create surplus value 
(i.e., are productive workers in the Marxist sense) -- researchers 
and data collectors are the miners of information production;  
programmers, the tool and die makers; computer operators, the 
forge hands; desktop publishers, the trim workers.

24. _Capital_, Volume I. p. 47.

25. See, e.g., Hindle and Lubar, 1986; Stallman, 1985; Wincor and 
Mandell, 1980. We are assuming here that the business is not the 
production of knowledge, like software companies. Their task is to 
sell the technology as widely as possible, while retaining 
ownership over the technology.

26. Knowledge-intensive companies like Microsoft (software) or 
Merck (pharmaceuticals) can be extremely profitable if they have a 
successful product.  Microsoft has two-and-a-half times the profit 
of Apple on a third of the sales. (_Business Week_, November 18, 
1991) Software companies are peculiar in that they carry virtually 
no fixed assets, and have little direct manufacturing costs.  A 
computer program that takes person-years to develop can be 
duplicated in a matter of seconds using a personal computer. 
_Soft*letter_, a software-publishing industry newsletter recently 
surveyed its readership regarding manufacturing costs. "The 
typical personal-computer-software company now spends 18 percent 
of its revenues on what accountants call 'cost-of-goods sold,' a 
category that includes product components, manufacturing 
operations, fulfillment, and shipping." The median unit cost of 
software products (disks and manuals) was $10.20; for products in 
the $300 - $999 price range, it was $20.98.  (Tarter, 1990)

27. "But in addition to the material wear and tear, a machine also 
undergoes what we may call a moral depreciation. It loses 
exchange-value, either by machines of the same sort being produced 
cheaper than it, or by better machines entering into competition 
with it. In both cases, be the machine ever so young and full of 
life, its value is no longer determined by the labour actually 
materialised in it, but by the labour-time requisite to reproduce 
either it or the better machine. It has, therefore, lost value 
more or less." (_Capital_, Volume I. p. 381.)

28. _Capital_, Volume I. p. 197.

29. "After making allowance [for wear and tear and auxiliary 
consumption like oil and coal, machines] each do their work 
gratuitously, just like the forces furnished by Nature without the 
help of man."  (_Capital_, Volume I. p. 365.)

30. _Capital_, Volume I. p. 365.

31. See, e.g., Forester, 1987.

32. Berliner, 1976.

33. From this, it follows that, in a knowledge-based economy, the 
production and distribution of knowledge becomes the leading 
sector of the economy, much like the railroad sector of the late 
19th century.

34. Forester, 1987.

35. "NeXTWORLD Interview: Alvin and Heidi Toffler", _NeXTWORLD_, 
Vol. 1, No. 2, March/April 1991.

36. "All Next's Factory Lacks Is Orders", _New York Times_. 
December 24, 1990. pg 23.

37. U.S. Bureau of Census, p. 413. According to the labor theory 
of value, as the value (exchange value, that is) of other 
commodities drop, so does the value of labor power, itself a 
commodity. Since price gravitates around exchange value, the price 
of labor power, i.e., wages, would drop as well.

38. "Help Not Wanted," _Business Week_, December 23, 1991.

39. Fusfeld, 1988.  See also Gandy, 1987 for a discussion of 
"information-rich" versus "information-poor."

40. "The rich are richer--and America may be the poorer," 
_Business Week_, November 18, 1991.

41. Miller, 1991.

42. "New York State's prison system has quietly imposed mandatory 
work policies, locking inmates who refuse work in their cells for 
23 hours a day and then blackballing them when they come up for 
parole... For [their labor] they are paid 60 cents a day at the 
start for a normal 40-hour workweek." "New York State Prisoners 
Work or Else." _New York Times_. Jan 27, 1992.

43. Peery, 1992.

44. "A Great Leap for software -- and Business", _Business Week_ 
editorial, September 30, 1991.

45. This process is no different than earlier efforts to break the 
power of skilled production workers. The automatic spinning mule,  
the cylinder textile printing machine, and the wool-combing 
machine all undermined the power of specific skilled textile 
trades. (Barsalla, p. 111).

46. "Thousands of Electronic Jobs Vanishing." _San Francisco 
Chronicle_. December 4, 1991.

47. "Both Tandem and Amdahl make large computers. That segment of 
the industry has suffered as customers move computing tasks to 
networks of inexpensive desktop machines." "Two Computer Makers 
Report First-Ever Losses." _San Francisco Chronicle_. January 24, 
1992.

48. Or Colorado or Texas, for that matter. Apple built its latest 
factory in Colorado, and located its technical support staff in 
Texas, because costs were cheaper there than Silicon Valley.
"American Firms Send Office Work Abroad to Use Cheaper Labor,"

49. _Wall Street Journal_, August 14, 1991.

50. Describing a recent arrangement between a Russian computer 
design team and U.S. computer manufacturer Sun Microsystems, the 
_New York Times_ reports: "The [Russian] team's full-time effort 
will come at an astoundingly low price for  Sun. Its members will 
be paid a little more than their current salaries of a few hundred 
dollars a year in American dollars... Top American computer 
designers sell their services for $100,000 a year  or more, but 
both Sun officials and Mr. Babayan said the Russians on the new 
team could not be paid that handsomely without engendering bitter 
feelings among their colleagues or causing inflation in the 
Russian economy... Other high-technology companies are searching 
for similar windfalls..." "Russian Computer Scientists Hired by 
American Company," _New York Times_. March 3, 1992.

51. The Boston 3 case is a recent example. The technical skills of 
three engineers, who were active in Irish support, was used 
against them by the prosecution. Knowledge, when combined with 
political conviction, now seems to be sufficient grounds for 
prosecution and conviction.

52. According to U.S. census figures some 7% of U.S. housing is 
vacant (U.S. Bureau of the Census).  San Francisco, with an 
estimated homeless population of 6,000 to 12,000,  has 22,000 
vacant housing units (_San Francisco Examiner_, April, 1991).

53. Sivanandan,  p. 8.

54. See, e.g., Hill, 1975.

55. Branscomb, 1986.


--------------------
REFERENCES

Barsalla, G. (1988). _The Evolution of Technology_. Cambridge: 
Cambridge University Press. 

Baumol, W. and A. Blinder (1979). _Economics: Principles and 
Policy_. New York: Harcourt Brace Jovanovich.

Berliner, J. (1976). "Prospects for Technological Progress." In J. 
Berliner (1988), _Soviet Industry from Stalin to Gorbachev_. 
Ithaca, NY: Cornell University Press.

Branscomb, A. W.  (1985). "Property Rights in Information." In B. 
Guile (ed.) _Information Technologies and Social Change_. 
Washington, DC: National Academy Press.

Cleveland, H. (1985). "The Twilight of Hierarchy." In B. Guile 
(ed.) _Information Technologies and Social Change_. Washington, 
DC: National Academy Press.

Cohen, S. and J. Sysman (1987). _Manufacturing Matters:The Myth of 
the Post-Industrial Society_. New York: Basic Books.

Costello, Michie, and Milne (1989). _Beyond the Casino Economy_. 
London: Verso.

Forester, T. (1987). _High Tech Society_. Cambridge, MA: The MIT 
Press.

Fusfeld, D. (1988). _Economics, Principles of Political Economy_. 
Boston: Scott Foresman and Company.

Gandy, O. (1988). "The Political Economy of Communications 
Competence." In V. Mosco and J. Wasko (eds.) _The Political 
Economy of Information_, Madison, WI: University of Wisconsin 
Press.

Goldhaber, M. (1985). _Reinventing Technologies_. Washington, DC: 
Institute for Policy Studies.

Hayes, D. (1989). _Behind the Silicon Curtain_. Boston:South End 
Press.

Hill, C. (1975). _The World Turned Upside Down_. New York: 
Penguin.

Hindle, B. and S. Lubar (1986)._ Engines of Change: The American 
Industrial Revolution 1790-1860_. Washington: Smithsonian 
Institution Press.

Kelly, K., "A bean-counter's best friend", _The Quality 
Imperative_, Business Week Special Issue, 1991.

Lash, S. (1991). "Disintegrating Firms", _Socialist Review_.

League for Programming Freedom (1990) "Against Software Patents." 
Cambridge, MA: LPF.

Marx, K. (1967). _Capital, Vol. I_. New York: International 
Publishers.

Marx, K. (1973). _Grundrisse_. New York: Vintage Books.

Marx, K. (1968). _Selected Works_. New York: International 
Publishers.

Miller, S. (1991). "The Electronic Chain Gang." Unpublished draft.

Peery, N. (1992). Talk presented in San Francisco, February 21, 
1992.

Sivanandan, A. (1990). "All that melts into air is solid: the 
hokum of New Times." _Race and Class_. London: Institute of Race 
Relations.

Tarter, J. (1990).  "Soft*letter's Manufacturing Survey", _Apple 
Direct, _April ,1990. Cupertino: Apple Computer, Inc.

Toffler, A. (1980). _The Third Wave_. New York: Bantam.

Toffler, A. (1983). _Previews and Premises_. New York: Bantam.

Toffler, A (1989). _Powershift_. New York: Bantam.

U.S. Bureau of the Census (1991). _Statistical Abstract of the 
United States: 1991_. Washington, DC.

Wincor, R. and I. Mandell (1980). _Copyright, Patents and 
Trademarks: The Protection of Intellectual Property_. Dobbs Ferry, 
NY: Oceana Publications, Inc.

(6/10/92/online version)


Jim Davis                     Michael Stack
jdav@well.sf.ca.us            stack@starnine.com