Skip Navigation
Angelo State University

Search Site

Information for:

Technology, Jobs, and Inequality: What is the Connection?

by Gary Chapman

In 1879, Henry George wrote in his classic book, Progress and Poverty. He wrote, “At the beginning of this marvelous era it was natural to expect, and it was expected, that labor saving inventions would lighten the toil and improve the condition of the laborer; that the enormous increase in the power of producing wealth would make real poverty a thing of the past.” However, noted George, “disappointment followed disappointment. From all parts of the civilized world come complaints of industrial depression … of want and suffering among the working classes.”

Shortly before George wrote his book, Karl Marx had observed the transformations wrought by technology in the modern factory. Marx wrote that technology:

“… transforms the worker’s operations more and more into mechanical operations, so that a at a certain point the mechanism can step into his place. Thus we can see directly how a particular form of labor is transferred from the worker to capital in the form of the machine and his own labor power is devalued as a result of this transposition. Hence we have the struggle of the worker against machinery.”(1)

These old controversies are being given new life in the “age of the smart machine,” a title of the noted book by Harvard Business School professor, Shoshana Zuboff.(2) Jacques Attali, minister of technology for the French government under Francois Mitterand, once said, “Machines are the new proletariat. The working class is being given its walking papers.”(3)

The state of the economy is among the most significant public issues for Americans, despite the present statistics of low unemployment, low inflation, and a booming stock market. Much of the anxiety in the U.S. middle class is rooted in trends of stagnant wages, flat or declining growth in the standard of living, and uncertain futures for young people. By now, everyone understands that the United States and the rest of the world are undergoing a major, epoch-making revolution in economics and production because of new technologies, a revolution every bit as historic and unsettling as the Industrial Revolution or the Agricultural Revolution of previous eras. Just as those revolutions remade the natural and human worlds, the current “information revolution” is reconfiguring the way we live, work, educate ourselves, and raise our children. Perhaps the main difference of the current revolution is the pace of change, which is so rapid it is disorienting and nerve-rattling for many, if not most, people. As the philosopher Alfred North Whitehead once observed, “The major advances in civilization are processes that all but wreck the societies in which they occur.”

Growing Inequality in the United States

The United States has always been a country based on the principles of freedom and equality, rather than on a uniform national or cultural heritage. Since the beginning of the U.S. as an experiment in republican democracy, freedom and equality have been two ideas that create friction with one another. The pursuit of economic freedom, of unfettered economic competition, it is sometimes argued, produces inequality. Conversely, policies that attempt to enforce economic equality constrain economic freedom. For much of the last 150 years, politics in the United States have been dominated by the debate over this dichotomy. This is still true today — the two-party system in the U.S. is essentially an institutional and ideological representation of this debate.

In the middle of this century, the United States and other industrialized nations reached a relatively stable consensus on the balance between freedom and equality, following a global depression, two world wars and the splitting off of a competing ideological bloc, the communist nations. This was the reign of Keynsian economics, a “social contract” between labor and capital, mediated by national governments, that produced unprecedented economic growth and a level of income equality never before seen in world history.

Between 1943 and the early 1960s, average wages for workers in the United States doubled in earning power, adjusted for inflation. This established a benchmark for the concept of the “American dream” of owning a home, a car, and working at a secure, stable job that would last a lifetime and eventually provide for an enjoyable retirement. It also cemented the expectation in the minds of Americans that each succeeding generation would be better off than its forebears, in material terms. Hard work, in other words, would pay off in tangible ways: improving material comfort, security, the lives of one’s children. This was in fact the experience of the broad middle class in the U.S., which became the backbone of the country’s political consensus.

The complaints of marginalized groups, such as racial minorities, were focused on how they were left out of this social contract. Martin Luther King, Jr., for example, believed that African-Americans should have a chance to experience the same kind of prosperity that white Americans were obviously enjoying in the 1960s. The fact that the civil rights movement did not challenge the legitimacy of the “American dream” produced support among most white Americans, and, in the mid-1960s, the middle class in the United States was feeling so secure that it supported the Johnson administration’s effort to eliminate poverty once and for all.

This experience ended in the early 1970s. Since about 1973, inequality has been increasing in the United States, and the U.S. now has the most unequal income distribution of any industrialized nation in the world. The immediate postwar picture is now reversed: the middle class is shrinking instead of expanding; wages for people without college education are falling, in real terms, instead of rising; wages for people with a college education are flat, on average, which means that a college degree is now a “defensive” protection against wage erosion, instead of a ticket to a higher standard of living.

The overwhelming share of new wealth created in the 1980s and 1990s has gone to the very top of society, principally to the top one percent of the wealthy. The poorest of the poor in the United States are worse off than the poor in any other industrialized country, and a growing proportion are children.

When I was younger, my elders used to claim that it was better to be poor in the U.S. than to be poor anywhere else. That is no longer true. In Italy, a country we used to think of as one of the poorest of Europe, the lowest 10 percent of male wage earners now make three times what the lowest 10 percent of their counterparts in the U.S. make.

The Sources of Inequality in the U.S.

There is no disagreement among economists in the United States that the country is experiencing growing inequality. However, there is no agreement on why this is happening.

In early 1995, economist Barry Bluestone published an essay in the magazine, The American Prospect, which was titled “The Inequality Express.” (4) Bluestone had originally titled his article “Murder on the Inequality Express,” meaning to link his argument with a plot device of Agatha Christie — I guess the editors of the magazine thought that putting homicide into the title was just a little too provocative.

What Bluestone argued was similar to the plot of Agatha Christie’s mystery, Murder on the Orient Express. Out of ten “suspects” as possible sources of growing inequality, all ten were guilty. Bluestone’s ten suspects on the “Inequality Express” are:

  1. Technology
  2. The service-based economy
  3. Deregulation
  4. Declining unionization
  5. Downsizing
  6. Winner-take-all labor markets
  7. Trade
  8. Capital mobility
  9. Immigration
  10. Trade deficits

Each of these items, as Bluestone noted in his essay, has its champions among economists. Groups of economists have staked out their “turf” by arguing that one suspect is guilty and the others are innocent or irrelevant. It’s only a few economists who have agreed that all of these suspects are guilty to some degree, and only a couple of economists have tried to sort out which is guiltier than others.

Richard Freeman and Lawrence Katz have tried to quantify the responsibility of factors, and, while their methodology is subjected to quite a bit of criticism, they are among the few who have attempted to assign blame in a systematic way. As you can see in Table 1, “technological change” has a rather striking range of responsibility according to Freeman and Katz; almost as much as trade and immigration. Whether technology is seven percent responsible, or 25 percent responsible, is a significant difference, with large implications for policy and remedies.

Table 1

Sources of Inequality:

Factors Responsible for the Increase in the
Male College/High School Wage Differential
During the 1980s
Technological Change 7%-25%
Deindustrialization 25%-33%
Deunionization 20%
Trade and immigration 5%-25%
Trade deficit 5%
Source: Richard B. Freeman and Lawrence F. Katz, “Rising Wage Inequality; The United States vs. Other Advanced Countries,” in Working Under Different Rules, edited by Richard B. Freeman, Beverly Hills: Russell Sage Publications, 1994.


What I’d like to argue — and this is my main theme — is that technology is very difficult to extract, as an independent factor, from any of the suspect sources of inequality. We can talk about technology as an independent variable when we look at capital investments, or at jobs requiring technological skills, or at numbers related to federal government codes for industrial sectors or product and job categories. However, this doesn’t really get at the true influence of technology on the trends contributing to inequality and job restructuring. So I’m going to offer another set of “suspects” as sources of inequality; suspects that are all artifacts of technological trends. Here are my broad categorical suspects:

  1. Automation
  2. Skill displacement and replacement
  3. Standardization
  4. Value-added hierarchies
  5. Intellectual property
  6. Data network globalization
  7. Re-engineering
  8. Disintermediation


Automation is the most obvious evidence of technological change in recent years. In August I toured the new BMW plant in Greenville, South Carolina, which is reportedly one of the most automated auto plants in the world, having just been built over the last four years. When the BMW Z-3 roadsters that are built there go through their painting process, which is a multi-step part of the production line, they are not touched by human hands. Everything is done by robots. A new Lexus plant in Japan is even more automated so that 60 workers produce about 400 Lexus cars per week. You may have heard the story about how a Saab production line in Sweden started making cars when there was no one on the line, and the cars were smashed up at the end of the line because there was no one there to drive them off.

A consequence of this is that we need far fewer auto workers than we needed 30 years ago. In the U.S. we produce more cars now than ever before, with about half as many workers as in 1978. In fact, worldwide, we’ve lost about a million auto jobs while we have a production capacity of ten million more cars, every year, than the market can absorb.

Similar things have happened to other so-called “traditional” industries, such as steel, textiles, shoes, and forestry. Nearly every manufacturing industry has undergone dramatic improvements in productivity over the past 20 years. Manufacturing employment in the U.S. hit a peak in the late 1940s and has been declining ever since, even while production is going up. Manufacturing employment is now between 15 percent and 16 percent of the total workforce in the U.S., and is steadily dropping as a proportion. Manufacturing production is undergoing the same trend line as agriculture, in which only about two percent to three percent of the workforce produces food for not only the entire population of the U.S. but a sizable export market, the largest in the country.

Automation is, of course, not limited to manufacturing. We’ve seen the appearance of automated teller machines, computerized telephone operators, automatic car washes, and so on. Automation of the service sector is only now beginning. Because advances in technology are startling everyone all the time, we’re not even sure which jobs can be automated and which cannot. We do know that there is a kind of “technological imperative” at work in the economy that works almost like a law. If a job can be replaced by a machine, it will be. In fact, some labor economists and sociologists argue that if a job can be replaced by a machine, it should be.

Within the past ten years, we’ve started to hear about completely worker-less factories, something known in the business community by the somewhat oblique euphemism “lights-out production.” This is no longer science fiction — within 25 to 50 years, we should see totally automated factories. We’re already seeing factories produce complicated and sophisticated products with only a handful of workers. For a year or two, every Macintosh computer coming out of the company’s plant in Fremont, California, was produced by about 50 line workers.

What this has meant, of course, is that an entire sector of the economy that once employed low to medium skilled workers is being wiped out by technological progress. Forty years ago, a person could make a decent living without a high school diploma and only the tools of his muscles and attention. That is no longer the case. And since 80 percent of our population and 75 percent of the workforce don’t have college degrees, you can see how much of workforce automation has affected. This is especially true among inner-city residents and the rural poor. William Julius Wilson, writing in his new book When Work Disappears, says, “For the first time in the 20th century, most adults in many inner-city ghetto neighborhoods are not working in a typical week.” (5) The jobs just aren’t there.

Still, this means that to run these automated factories, companies need skilled workers such as robotics technicians, programmers, process control engineers, and so on. People with these skills are scarce in our economy, so they are paid well. When low and medium skilled workers are put out of work by technology, and higher skilled workers are rewarded for their expertise, that increases the wage gap. In the 1970s, the wage difference between a high school graduate working in manufacturing and a college graduate working in the same sector was a factor of two; now it’s a factor of three and it’s going up.(6)

Skill Displacement and Replacement

In addition to the restructuring of the workplace because of automation, worker skills are also changing. Throughout most of this century, a major employment category of women was “secretary.” Since the 1980s we’ve lost 600,000 secretarial jobs in the U.S., and the job category has dropped, dramatically, in status. It was once viewed as a respectable white-collar job for women, even something denoting rank and sophistication, a solid middle class job. Now it’s commonly viewed as a dead-end job, without upward mobility, low-paying and constantly threatened by pay cuts and downsizing.

The main reason for this, of course, has been the introduction of the personal computer, which has forced most white collar managers to do their own typing and filing. The PC has also lowered the quality of work for clerical workers, because they are now typically left out of the stream of managerial decision-making and relegated to word processing, photocopying, taking phone calls, and filing paper documents. Secretaries were once considered the unspoken rulers of an office — they knew where everything was, what everyone was doing, and what the real truth was in a company. Now they are commonly temporary employees hired from a company like Manpower, Inc., which has become the nation’s largest employer. All they really need to know is how to drive a word processing application.

Or look at grocery check-out clerks. They were once a source of community conversation and networking, and they knew the prices of the goods you brought them. Now they’re essentially humanoid robotic arms, scanning bar codes on products. In the near future, holographic lasers will scan an entire shopping cart at once, and you’ll pay with a smart card, so there will be no need for check-out clerks at all. All that will be left will be someone to help you take your groceries to your car, a job which, at my local grocery store, is filled by modestly competent people who are mentally retarded.

This skill displacement is related to automation, but it has different consequences. Instead of jobs merely being eliminated, they are changed to require less skill, less familiarity with the task, and, consequently, the jobs require lesser skilled people. Training costs are lowered, wages are lowered, and the high turnover rates that these jobs typically have are less costly to employers. McDonald’s employees can be trained in an afternoon, because the technology and the division of labor are so intertwined that the employee is usually only required to push a few buttons and keep things clean.

Labor specialists call this the “dumbing down” of jobs made possible by putting intelligence into machines, instead of requiring people to develop skills. This trend is exhibited in workplaces like those for telemarketers, who fill out computerized forms, read computerized scripts, and whose work is paced by the machines they sit in front of. Directory assistance operators used to talk to you while they tried to determine what you wanted; now they just type in what you ask for and push a button to have a computerized voice read you the number.

All of this makes human beings tools of their tools instead of the other way around — it makes people essentially adjuncts of technology, such as the eyes, ears, voices, arms, and legs of machines. This “dumbing down” process lowers wages, increases the sensitivity of these jobs to further automation, and survives on a huge number of unskilled people who have no other employment options. This quite obviously increases the trend toward inequality.


By standardization, I mean the universal trend toward economies of scale, especially using standard technologies which can structure work not only in one workplace, but in many at the same time. Take a business like Blockbuster Video, for example, which reportedly has over 4,000 outlets in the United States. All Blockbuster stores use the same computer system, the same inventory procedures, the same methods for registering customers. Their signage, their ads, their promotions, their employee uniforms are all the same or very nearly identical. If Blockbuster management decides to change something about its computer system, it will affect thousands of stores and tens of thousands of employees simultaneously. While Blockbuster may invest far more in computer services than a local “Mom and Pop” video store, the per-store investment is far smaller — that’s an economy of scale made possible by both size and technology.

This is basically how Wal-Mart became the biggest retailer in the world. It standardized its inventory system, linked stores through direct broadcast satellites, developed a highly integrated point-of-sale and inventory management system, and pulled inventory data to its Arkansas headquarters every single day. The system was so sensitive to sales trends that Wal-Mart is famous for changing its prices on a daily basis, either to move inventory by lowering prices, or to take advantage of popular items by jacking up prices. Small increments in price for tens of thousands of items in each store, adjusted by computer programs that track this inventory every 24 hours on a world-wide basis, produce aggregate profits that no other retailer can match. So Wal-Mart has knocked Sears into a tailspin, put K-Mart into bankruptcy, and buried small local stores on Main Streets throughout the U.S. Wal-Mart runs like a machine, and that means that most of its employees are doing nothing but oiling this machine by running the cash registers, changing the bar codes, stocking merchandise, and sweeping the floors. Wal-Mart is the largest employer of on-site personnel in the U.S. — with about 628,000 employees — and the vast majority of these people make just above the minimum wage.

Go to just about any shopping mall in the United States and you will see the same stores: the Gap, Kinney Shoes, The Limited, Sunglasses Hut, etc. These stores are successful because of a combination of standardization and technological systems that take advantage of economies of scale. The retailers are typically mated to manufacturers who practice similar methods. Some clothing stores are even starting to experiment with “custom” tailoring. Levi’s, for example, already has so-called “just-in-time” production systems that minimize unsold inventories because they’re so sensitive to the sales of Levi’s products in retail outlets. The next step is “right-on-time” systems that will come close to producing custom-tailored clothes at mall stores: step into a computer that will take your measurements, the data will be sent to an automated production facility, and “your” jeans will come back in a couple of days.

The synergy between economies of scale and standardized technological systems for production and sales gives huge advantages to large employers. That’s why we’re seeing concentration in almost every industry, a dwindling of competitors, and this means that money that is drawn from consumers goes to the bank accounts of a smaller proportion of owners, once again increasing inequality and concentrating wealth.

Value-Added Hierarchies

A phenomenon related to standardization is what I call “value-added hierarchies.” Because of the automation of production, the standardization of technologies that help manage production and sales, and the increasing uniformity of the machines used to accomplish these tasks, we’re building hierarchies of people who add value in accumulating degrees. Let me try to explain this.

If you are a software programmer, for example, you don’t really make anything. What you do is allow people to do other tasks more efficiently, and the variety of tasks they undertake using your software will be as varied as the number of copies you sell. In effect you try to generalize the properties of work and turn those into algorithms that become features of your program. If you are successful at doing this, you are adding value to the task your customers undertake — that’s why they pay you for your product. The more tasks that you are able to generalize effectively, and code into your software as conceptual frameworks, the more customers you will attract and the more money you will make. That’s why Netscape makes more money than a company that makes computerized Rolodexes — you can do more things with Netscape Navigator.

In effect, Netscape and similar companies — most especially, Microsoft — are sitting at the top end of an immense hierarchy of value, with each layer of this hierarchy adding a little more value depending on its contribution to the solution of a problem. The clerk at Blockbuster adds a little value to a transaction; the guy who designed Blockbuster’s inventory software, which is used in all their stores, adds enormously more value to Blockbuster.

This is not essentially different than what Henry Ford did during his heyday; or J. P. Morgan, or Rockefeller, or their top engineers, or finance experts. What is different today is that we have tightly coupled and vastly expanded networks of value. Ford sold cars — his customers were car and truck buyers. Bill Gates sells something else — the structuring of tasks with his technology. Because his chosen technology is the “universal machine,” he sits on top of the ultimate pyramid of value.

Hierarchies of value are also related to one of Bluestone’s suspects on the inequality express — “winner-take-all”” markets. These are described in a book called The Winner-Take-All Society, by economists Robert H. Frank and Philip J. Cook.(7) I won’t describe this phenomenon in much detail, but in its essence, the “winner-take-all” argument says that our current labor market has a built-in irrationality in that the rewards to winners are far and away more than the rewards to runners-up, not to mention everyone else, even in a given profession. The lottery-like rewards to winners distorts the choices of people when they choose their professions, which in turn distorts labor markets even further. This is aided and abetted by technology. For example, Michael Jordan would be paid far less than he currently makes without television. Movie stars would make less without VCRs, satellite broadcast of movies, TV talk shows, and so on. Bill Gates would be worth far less if there were more varieties of computer operating systems available, as there once were.

There is a kind of “cascade” effect, throughout our economic system, as winners of value “cross over” into ancillary markets and secure product “tie-ins,” brand-name alliances, and so on. Such tie-ins are exemplified, for example, by bringing flashy and zingy ads for shoes to TV broadcasts of Chicago Bulls games, or when Microsoft employed the Rolling Stones to help sell Windows 95. Again, the technology is integral, since these tie-ins are essentially conceptual artifacts, not real things. The only way they become “real” is through technological dissemination.

There is, once again, a synergy between standardization and hierarchies of value. If economies were not linked by technological systems, if the global cascade effect did not happen, if standardization was limited to offices instead of covering entire countries or regions, we would have multiple hierarchies of value, shorter pyramids, and less inequality. Because of these trends, tinkering with technology at the top of the pyramid — such as at Microsoft — affects the value scale around the world. Because of this, money flows to the top, and inequality increases.

Intellectual Property

The phrase “intellectual property” is regarded by some people as an oxymoron, and its widespread use in our society is relatively recent. It is an increasingly important concept in the digital economy, especially as the most lucrative “products” of the economy become disembodied, concept-dependent goods for sale; products and services such as ideas, music, images, advice, or systems of organization.

The subject of intellectual property is nearly endless in its scope, and very controversial. I cannot do it justice here, nor even touch on the basic controversies. I do want to stress, though, that intellectual property’s technological contribution to inequality is, once again, linked to other trends — specifically standardization and hierarchies of value.

It is not enough that automation, standardization, recasting of skills, and new hierarchies of value should merely happen as a result of technological developments. Intellectual property adds the critical feature that someone should own the elements of these trends. This is how hierarchies of value become something more than just an abstract concept — people make money by owning the concepts that add value. So if you invent the word processor that 50 percent of computer users work with every day, you not only have a formal place in a hierarchy of value, but by virtue of intellectual property law, you get rich too. The more people you affect down a chain of utility, typically, the more money you make. Because of intellectual property law, you can come up with an idea, and if it ranks you on a hierarchy of value, you can cash it in.

Just about ten days ago it was reported that a graduate student in computer science at the University of Washington sold his Ph.D. dissertation, as yet incomplete, to America Online for $1.4 million, plus additional stock options. He bought a $600,000 house and tore it down to build a bigger one. While this sort of opportunity fits the American ideal that riches can be had if you have a good idea and work hard, it also means that people who are wired into a hierarchy of value through connections, recognized talent, and intellectual breakthroughs are increasingly rewarded in disproportion to their real value. One could argue, for example, that any grade school teacher could have more influence than this young computer scientist, but a grade school teacher will probably not make $1.4 million in her entire career. The combination of intellectual property, standardization, and hierarchies of value exacerbate the trends toward a “winner-take-all” society, increasing inequality.

Data Network Globalization

If you are in Washington, D.C., and you call directory assistance, you will speak with an operator, not in Washington, but in West Virginia where wages are about two-thirds of what they are in the District of Columbia. If you work for a large law firm in Los Angeles, your word processing documents will probably be done by a typist in Hong Kong or Taiwan, and if you’re in New York they’ll be shipped back and forth overnight to Bermuda or Ireland. If you are in Italy and you call an 800 number to get information about visiting New Mexico, you’ll probably speak to one of the service operators for the tourist bureau of the State of New Mexico, and he will probably be in the New Mexico state prison, as an inmate.

Globalization of the economy is one of the most frequently mentioned causes of economic changes in the U.S., because it has put American workers in competition with every low-wage labor market in the world, from Central America to Southeast Asia to Eastern Europe. When Mexican workers can do the work that Americans once did for about $3.50 per day — which is what some Mexicans are making in U.S. company plants in northern Mexico — or when Salvadoran workers are making about 35 cents an hour making clothes or textiles, American workers are going to lose out. And those who still have jobs will see their wages fall too.

What is less frequently mentioned about globalization is that none of this would be possible without the technology that we otherwise celebrate. International data networks make global production feasible. Satellite transmissions of inventory reports, production quotas, e-mail, blueprints and specifications make possible the radical separation of management, design and production, which is what makes global enterprise possible. My father, for example, is a retired mechanical engineer. He lives in Solvang, California, a sleepy little town north of Santa Barbara. For some years, he consulted with a U.S. company that makes a blood-filtering machine. He designed the mechanical components on his $40,000 Hewlett Packard CAD computer in his office at home, in a building behind his house. One of the factories making this blood machine was in Kyoto, Japan. My father’s computer was hooked to the computer-controlled milling machines of the Japanese factory by modem. When they were developing the prototype parts, to be used for making automated tools that would actually make the blood machine, they would go through several iterations of the part every day. The blueprints were tweaked in California, then the parts were milled a few minutes later in Kyoto, Japan. The Kyoto engineers could instantly tell my father what worked and what didn’t work.

Global data networks also intensify and broaden the international transfer and manipulation of finance capital. Over a trillion dollars per day moves through cyberspace between banks, investment firms, traders, and speculators. Governments have long since lost control over these capital markets which can make millions for individual investors in a week or even days. New York City investor, George Soros, makes and loses hundreds of millions of dollars on single days by betting for or against national currencies in so-called “hedge funds.” Some people believe that a single hedge bet by Soros caused the British pound sterling to drop seven percent in one day in 1994.

Computer-driven stock trading is now the standard method of most Wall Street firms where stocks are bought and sold based on the results of staggeringly complex financial formulae processed by extremely powerful computers. Automatic computer trading is thought to have been responsible for an otherwise inexplicable acceleration in the drop of the stock market in October of 1987, on the day known as “Black Monday,” when the U.S. market lost over $6 trillion in value in six hours.

The mobility of capital and jobs made possible by data networks is only beginning. Already we’re seeing an explosion in growth in communities once thought beyond the reach of high finance and commerce, such as in Durango, Colorado, and Boise, Idaho, and in remote parts of Montana. The “winners” in the digital economy are not only capable of commanding high salaries but of living wherever they want and “telecommuting” via computers, modems, and networks. This concentration of the affluent in highly desirable communities skyrockets land and home values in those places, often displacing people who have lived there for generations, and it also draws money away from places that need investment capital, business, jobs, and skilled citizens.

We’re also beginning to hear talk of a new kind of escape from the problems of an unequal society. “Electronic commerce” is in its infancy, but if it takes off, people will run businesses over the internet, communicate with friends, do their own banking and investing, order their groceries, play games, all without leaving a room. Accompanying these predictions is a certain fashion of “cyber-libertarianism,” especially popular among computer industry professionals, that denies any obligation to the nation or to any geographic community. There is a thread of Social Darwinism in such pronouncements (found, for example, in Wired magazine, and frequently in various on-line forums on the internet). Inequality, in this view, is not a social problem, but something unavoidable and even just — a metric of merit.


The term “re-engineering” is trendy among business consultants and managers. Critics sometimes say that it is synonymous with “downsizing,” an equally oblique euphemism for layoffs, firings, and large-scale elimination of jobs. Proponents of re-engineering insist that it’s a way of reconfiguring companies to take advantage of the capabilities of information technologies.

“Re-engineering” arose as both a concept and a management fad at about the same time. A distillation of the work of many different business advisors, including Zuboff, Peter Drucker, and especially the late Edward Deming, “re-engineering” became a way for companies to look at their core activities, then rebuild their organizational structures around technology rather than merely add technology as an “applique” over existing organizations and procedures.

However, a byproduct re-engineering has been massive layoffs in the broad category of white-collar work known as “middle management.” Company computer networks, information systems, and reconfigured, “flattened” levels of authority within companies have made millions of middle-managers superfluous. Since 1988, the U.S. economy has lost over one million middle management, white-collar jobs, and it appears these jobs are gone for good. The psychological shock of this shift has been severe among those who have been its victims, because many of these people have been middle-aged, with children in college, and with expectations of lifelong employment and generous pensions. The majority of them have been able to find work only at lower levels of pay and with fewer benefits, if any.

Re-engineering is not over. It has really only been significant in very large firms that had personnel overhead dating from managerial concepts of the 1950s and 60s, like IBM, AT&T, Kodak, Xerox, General Motors, and other big, traditional, “blue-chip” firms. Re-engineering has not yet happened in about 80 percent of the businesses in the United States. As technology becomes more capable and more sophisticated, and as firms restructure to remain competitive, we can expect that even more middle management jobs will be eliminated, probably permanently.

Another feature of re-engineering that is related to technology is that it makes workers be entrepreneurial. White-collar workers are expected to keep up with technology, to sell themselves, and, increasingly, to assume that jobs will be temporary. Of course, when you are 50 years old and have spent 25 years with one company, and you are suddenly on the job market with 25 year-olds who have spent 14 hours a day learning the very latest in technology, your chances of competing are slim. Moreover, younger workers with high-demand skills are willing to work for less money than “downsized” middle-aged men with mortgages, college expenses, and worries about retirement. Because technology is changing so rapidly, it is increasingly difficult for anyone to keep up with it.

Re-engineering contributes to inequality in two ways. First, it eliminates certain kinds of jobs that are not replaceable; more than two-thirds of the past five years’ “downsized” executives wound up in jobs paying less than what they made before. Second, it creates a two-tier system within firms, with less upward mobility inside the company. There are upper level, “creative” managers who run the firm, there is support staff, and there are few people in-between. Pay scale differences widen. Now, for example, CEOs of companies, on average, make 180 times what their lowest paid employees make. That figure was 50 in 1980, and in Japan it is still only about 25 or 30.


“Disintermediation” is one of those terms of academics that is by the very sound of it totally opaque. What does it mean? Disintermediation means that there is a process under way in society, because of new technologies, that allows the elimination of a vast sector of employment and business, a sector typically called “middle men.”

One example should be familiar. People can now buy home satellite dishes and order movies that they might have otherwise rented on a videocassette from a local video store. That’s “disintermediation” — the video store loses a sale because the satellite dish owner doesn’t need the store or its employees to get that particular movie. Now, imagine that you can order any movie you want over the internet, which some people say is a sure thing in the future. When that happens there will be no need for video stores at all, or video store clerks, or the people who sell things to video stores. You may be able to buy other things over the internet too, and that will cut out the stores that would have sold those goods to you. Someday you might eventually be able to program your own little software robot, or agent — something that already has a name, a “bot” — and send it out on the internet looking for a good deal on, say, a new washing machine. The best price and features ratio may be from a manufacturer in northern China, or upper Uzbekistan. You’ll order the washing machine, pay for it on-line, and it will be delivered to your house, perhaps by a robot delivery vehicle. No more salesman, no more store, no more ad in the newspaper. That’s “disintermediation.”

This is not something that will only happen in the future, someday. Automatic teller machines are examples of disintermediation; so are book and record clubs, mail-order sales, reading articles on the internet, buying something on the Home Shopping Network, and so on. This is happening more and more and people who are successful at disintermediation are making a lot of money. To read a gushing prediction about the importance of disintermediation in the future, read Bill Gates’ best-selling book, The Road Ahead.

Disintermediation contributes to inequality because it eliminates jobs, wipes out categories of skilled labor, and it once again polarizes earning potential at two ends of the extremes — a large number of people who will be telemarketers, internet content designers and programmers, and various technicians at one end, and “creative” symbolic analysts on the other end. In addition, disintermediation fosters centralization. If people were to do all their banking on machines, for example, the bank itself could be anywhere. If everyone bought their movies over the internet, the movie houses would go out of business, and Hollywood would get even bigger than it is now. Money, instead of being widely circulated in a local economy, would concentrate in centers where “disintermediated” companies collect.

Technology Policy Options and Inequality

There are a number of proposals on the table for helping ameliorate economic inequality in the United States, although no one has suggested that inequality can be eliminated completely. Some proposals, such as for more education, are well-meaning but in fact do little to solve inequality. Given other trends, more equal access to education, for example, simply makes marginal advantages in education more valuable and leaves inequality unchanged. In fact, in the U.S., education has become more equally accessible over the past 40 years. It is set into the context of a dispersed meritocracy. American education is not designed to produce a more equal society.

The most direct way of attenuating inequality has the least amount of political support: transfer payments, or deliberate redistribution of wealth. Tax policy for the past 15 years has instead fostered accumulation of wealth by the affluent, instead of redistributing it to the needy. Earned income tax credits have bipartisan support, but only at very low levels that do little to change the order of things. Earned income tax credits are also a form of subsidizing low wages paid by employers, which dampens incentives for raising wage rates. Public works measures, meaning public sector jobs as a “last resort” of employment, are supported by some economists, but very few politicians or voters these days. As a wage measure, only this year’s passage of the minimum wage bill, which will raise the minimum wage over the next two years, will have any significant effect on the incomes of the working poor. Even that is likely to be offset by modest inflation, lost jobs, or the effects of a recession, assuming there is another downturn in our future.

I want to focus on some ideas related to technology. Most of these are not part of the nation’s political debate yet, although they are well-developed proposals in other countries — particularly in Europe. The ideas I’d like to review also have a growing constituency here, in the United States, where there are various advocacy groups and community organizations that are almost never mentioned in the press or in academic journals.

Here are my proposals for technology-based ways to help attenuate inequality:

  1. Skill-based automation
  2. Participatory design
  3. Reformed unionization and the High-performance workplace
  4. Local economies of scale
  5. Enhanced access to the internet
  6. Democratic technology policy
  7. Targeted public technology investments

Skill-Based Automation

Skill-based automation is a term that comes from Germany and Scandinavia, and it is largely unknown in the United States. It’s a pretty simple concept. It means developing ways to automate the workplace that preserve and enhance skills instead of replacing them. The idea is best illustrated in the fact that, in the United States, we have an entire class of software products called “expert systems,” which in Scandinavia are called “systems for experts.” Turning this phrase around implies a vast set of assumptions about the way that machines and people should work together.

In the 1980s most research in artificial intelligence was funded by the Department of Defense. This work, in the early 1980s, concentrated on systems of command and planning, so-called “battle management systems” for example. In other words, the military was attempting to automate decision-making that was the most complex, the most context-dependent, and the most intuitive and skill-based of any. This approach didn’t work. The military has now switched to providing information systems that give commanders data they need to make decisions. This approach works quite well, and it is mirrored in the business world by a switch from “expert systems” to “executive information systems.”

The same approach has yet to reach planners who work with shop-floor employees, or medium to high skilled production workers. When file clerks or telephone operators are replaced by automated systems, they are shown the door, not retrained to be higher-skilled operators of machines that help them in their jobs. In most U.S. business schools, labor is still treated as a “factor of production,” or a “cost center” that can be cut — that’s why we see company stock prices go up when the company announces layoffs.

Skill-based automation is a design methodology that attempts to look at the way a job is done, who is doing the job, and what sort of automated systems would help these people do their jobs better, faster, more efficiently and even more pleasurably. An excellent videotape about this idea, called Computers in Context, produced by California Newsreel, documents the experience of some typesetters in Norway who were about to lose their jobs because of “desktop publishing” systems. A union-funded research project developed an alternative system that not only saved the typesetters’ jobs, but cut costs and streamlined the production process.

Skill-based automation is an idea still in its infancy in the United States, but it has its champions. Frank Emspak, a professor at the School for Workers at the University of Wisconsin in Madison, has developed a skill-based automation program for production cabinet makers in the Midwest. The International Machinists’ Union and other trade unions have also investigated skill-based automation concepts. More work in this direction could be done, and the federal government could support such work through its technology investment programs.

Participatory Design

Participatory design is another idea imported from Scandinavia. The largest participatory design research program in the world is called the Utopia Project, based at the University of Aarhus in Denmark. Participatory design is, like skill-based automation, a methodology for technologists. It attempts to involve the user of technology in the technology’s design, especially, in computer systems, and in the human-computer interface. The motto of participatory design advocates is “not for the user, not by the user, but with the user.”

Perhaps the most significant example of participatory design in the United States has been the recent boom in innovation in technologies that serve disabled people. The disability rights movement successfully brought complaints about access, dignity, and other social values to considerations in the design of buildings, computers, elevators, doors, sidewalks, parking lots, and other daily technologies. What many of us now take for granted as innovations that serve disabled people are in fact the result of an intense and deliberate process of participatory design that has engaged disabled users in new designs for common technologies. There are even think-tanks set up for this purpose, such as the Massachusetts Assistive Technology Project in Boston.

A new job category that is increasing in popularity is “office anthropologist.” Companies are hiring people trained in anthropology to study the way offices are organized, both socially and physically, in order to optimize efficiency and morale. The Steelcase Company, which makes office furniture, has a large project in participatory design that brings together clerical workers, anthropologists, office planners, architects, and systems engineers to study how to make offices more efficient and comfortable. American Airlines has also sponsored a large participatory design project in the redesign of its SABRE on-line reservations system, building a team of reservation clerks, software engineers, managers, and customers.

Participatory design, like skill-based automation, is a way to enhance skills, prevent skill displacement, and increase productivity at the same time. All of that should produce higher value and higher wages. No one who has practiced participatory design reports an instance of design participants lowering the skills required to work with the system they are developing. Studies have also shown that information technologies work best when they are used by people who feel that the systems are serving them, instead of feeling like they’re chained to a pre-defined work process or a rigid computer system.

Reformed Unionization and the High-Performance Workplace

Skill-based automation and participatory design work best in unionized workplaces, at least when the union is committed to and supportive of such initiatives. Independent worker representation, through genuine trade unions, is the best means to produce an “empowered” workforce that feels confident about its role in the workplace and the technology it uses on the job.

Trade unions have been on hard times over the past 20 years in the United States. Trade union membership now stands at its lowest point in the 20th century, about 14 percent of all workers and only about 10 percent of private-sector workers. Trade unions have low public approval because of a history of scandals, ties to organized crime, and regrettable leadership. Unions are also the target of a well-funded campaign of bad press, union busting, and lockouts by corporations opposed to unions, especially in high tech companies, and especially in the South and Southwest. High tech executives claim that the union traditions of work rules and hierarchies of tasks are obsolete in the fast-moving information economy in which workplace flexibility and “flattened” structures of authority are imperative. Some trade union leaders are trying to bridge this gap with a new role for trade unions — building the “high performance workplace.” This is an immense subject that I can only touch on here, but the chief advocates of this new approach are my colleague, Ray Marshall, at the LBJ School of Public Affairs, and the researchers at the Work and Technology Institute in Washington, D.C., which is run by Dr. Brian Turner.

The “high performance workplace” combines most of the recommendations I have described so far with deliberate social reorganization of the workplace for maximum performance and value. The proposals include worker-managed teams, individual worker empowerment, an end to Taylorism, participatory technology design, skill-based automation, and collective bargaining agreements that include technology planning committees with worker membership. Most of these elements are already part of the labor-management agreements in Europe, particularly in Germany and Scandinavia.

The goal of the “high performance workplace” is to pursue a “high wage, high skill” strategy for American workers, instead of making them compete against unskilled workers in other countries who will settle for 10 percent or less of the lowest American wage. Because the “high performance workplace” paradigm is based on increasing skills and productivity, it is necessarily integrated with a different approach to deploying technologies in the workplace.

Local Economies of Scale

In contrast to the global economies of scale that are often pursued by giant multinational corporations, the United States could — as a matter of policy — help promote local and regional economies of scale that would help keep money circulating among a greater number of people. And this can be done with certain kinds of technological investments.

The first kind of technological investment that would help promote local economies of scale is related to environmental protection. Conservation, recycling, independent consumer control over sources of energy, and various other kinds of local innovations for the environment, can help build local businesses and keep money in a community. The alternative is actually what we’re pursuing now — deregulation, increased dependence on fossil fuels, national or international energy cartels, all of which concentrate money at the top of a pyramid of value.

Another kind of investment that would help build local economies of scale with technology involves community computer networks. Community computer networks, with public access stations, pooled community computing centers, low-cost access to community on-line resources, help with on-line commercial start-ups are an alternative to a global information network dominated or controlled by giant media and telecommunications companies. Again, unfortunately, our public policy is moving in exactly the opposite direction, favoring immense companies with a lot of political clout. We are neglecting the “town hall” kind of computer network in favor of a giant “shopping mall” model.

The goal of fostering local economies of scale is to multiply and shorten the pyramids of value I discussed earlier. Instead of money flowing to a handful of powerful and remote corporations, local economies of scale could keep money circulating in a community and result in greater opportunities for participation in money-making activities. This should be accompanied by an aggressive policy of supporting community development banks, Grameen-style lending clubs, management assistance, and infrastructure development grants.

Enhanced Access to the internet

About 12 percent to 15 percent of the U.S. population uses the internet today, and that number is doubling every year. Internet users are concentrated in the upper levels of income strata, and they are overwhelmingly white men with upper-middle class salaries. The median income of an adult internet user in the U.S. is about $67,000 per year, nearly double the median income for a U.S. family of four. At the bottom rungs of society, the internet is nonexistent. Only about four percent of adults with incomes less than $15,000 per year use the internet, and that figure is almost entirely students who use the internet for their schoolwork. Among the 30 million or so people who fit into that income category, the internet is largely invisible and inaccessible.

Today, roughly two-thirds of jobs in the U.S. require use of a computer on a daily basis, and that proportion will probably go up to about 75 percent by the early years of the next century. The lack of routine and on-demand access to a computer and the internet has become a serious impediment to success in the job market. Yet among people who need to improve their skills the most, computers and network access are scarce to nonexistent.

There are modest government programs to help solve this problem, such as the federal Department of Commerce’s Technology and Information Infrastructure Assistance Program (TIIAP). However, these cannot provide enough money at current levels of funding to make much of a dent in the disparity we already see today. Current TIIAP funding is less than $25 million per year for the entire United States.

What is required is a national, public commitment to providing universal access to the internet — something that will conceivably cost a billion dollars or more over several years. President Clinton has asked for $2 billion to wire every classroom in the U.S. to the internet, but this is only a first step. The internet needs to be accessible in communities the way we provide pay telephones, public transit, and public recreation facilities. So far there is not a consensus for such investments the way there is for public transit or public parks and recreation centers.

We should also change the character of “access,” which is a term currently dependent on concepts derived from the telephone era. When the technology under discussion is an interactive computer linked to a global network, the question immediately arises, “access to what?” It does no good, in terms of dealing with inequality, to simply provide the opportunity for network access if the technology required to hook up the network is too expensive, or if the skills that are required are unavailable, or if the content on the network carries a price tag that adds up to an amount that only a few people can afford. So access to the internet cannot mean what it has meant for access to the telephone network, at least if we want equitable access. The technical term of “universal service,” which essentially means equal opportunity to connect to the network, must be accompanied by “enhanced equitable access,” which should entail programs of community networking, public access stations, local content development, and various other means of making the network “friendly” and accessible for low-income citizens.

Democratic Technology Policy

Throughout the decades of the Cold War, U.S. science and technology policy was not only dominated by military priorities, but also by a model of decision-making that relied exclusively on experts and the “high priests” of academia, think-tanks, the military-industrial complex, and government. Not surprisingly, their recommendations for technology investments tended to reinforce the positions and status of major corporations, elite schools, and government agencies.

Now that the Cold War is over, there is an opportunity to reform this model of policy-making for science and technology and to make it more democratic. There are two challenges, however. One is to make scientists, engineers, and policy-makers understand that ordinary, non-expert people deserve to be heard on how technology is developed in our society; the other challenge is to develop alternative vehicles for meaningful citizen participation. These things are not common in U.S. public life.

Studies by the Kettering Foundation and the Public Agenda Foundation have found that ordinary, non-expert citizens not only understand controversies involving science and technology, but that they often come to very similar conclusions as experts. Given enough information from non-partisan sources, or partisan information balanced by opposing views presented fairly, citizens can make informed choices about how technology should serve human and natural needs. This, in other words, is not the problem. The problem is getting policy-makers, and even more so, scientists and engineers, to believe that citizens have a proper and useful role in policy-making for scientific and technological investments. That is a political problem, not a scientific or technical problem.

There are many models available for increasing citizen participation in scientific and technical decision-making, such as Denmark’s “consensus councils,” “science juries,” the focus groups used by Public Agenda, or the “deliberative opinion polls” developed by University of Texas political scientist, James Fishkin. From 1993 to 1995, with support from the National Science Foundation, The 21st Century Project at the LBJ School of Public Affairs at the University of Texas explored the development of a National Citizens’ Forum on Science and Technology, building on a recommendation of the Carnegie Commission on Science, Technology, and Government. This forum would probably use a combination of techniques, if funding for such an initiative were available. Unfortunately, there is very little public pressure for creating such a forum, and consequently little action on the part of policy-makers to change the status quo. Nevertheless, a democratized science and technology policy could foster more diversity in government policies, greater opportunities for low-income neighborhood investments, and potentially different kinds of technology investments — away from military programs which still dominate federal budgets, for example, and toward environmental, worker-friendly, and community improvement technologies.

Targeted Public Technology Investments

When Bill Clinton was elected to the White House in 1992, one of his first documents spelling out policy initiatives was a “white paper” on technology policy, titled “Technology for America’s Growth.” The ideas presented in this paper were linked to Clinton’s $60 billion “infrastructure investment” plan, which was dead by the end of 1993, killed by Congress’ concern over federal budget deficits. Now, after both Congress and the President have committed to balancing the budget in seven years, there is little chance that we’ll see any proposals for large federal investment programs. In the current campaign, President Clinton’s ideas for what he will do in his second term are comparatively modest — the largest is probably his internet in the classroom appeal pegged at $2 billion.

In my view, there is still room in a balanced budget for targeted technology investments to help lower inequality, if only we would shift our spending priorities. The U.S. spends more on defense than the next eight countries combined, including nearly $640 billion per year on defense research and development. Some portion of that money could be redirected to civilian priorities — just $3 billion per year, for example, would double the size of the National Science Foundation; about the same amount of money we spend on “Star Wars” research, or on our nuclear weapons laboratories which are no longer researching nuclear weapons.


We need to develop a consensus about what this country should invest in over the next 50 years in the same way we developed a consensus for science and technology funding after World War II. The 21st Century Project has proposed three areas of investment: environmental technologies; high performance workplace technologies; and equitable information infrastructure, skills, and access. With targeted investments, we could build a “green” economy, a “high-skill, high-wage” economy, and one that would bring low-cost information access to every citizen. This would make the country more productive, cleaner, healthier, and more literate and participatory. It would help develop both democratic citizenship and high-skill workers.

The alternative is for the United States to become more like Brazil, or Mexico, with a tiny rich portion of the population, surrounded by hordes of poor, living behind walls topped by broken bottles and guarded by armed security forces. To a certain extent we’re seeing that happen already, and it’s a frightening sight. Middle-class whites are escaping the cities to live in “gated communities” with homogeneous, conservative neighbors. There is, of course, a wave of xenophobia sweeping the country, as well as callousness about the prospects of the poor. A new elite of my generation is gradually accepting the ideas of Social Darwinism. Along that path is ruin and cultural collapse, the end of democracy.

So the message of this talk is that technology is not something to be viewed simply as a mechanical artifact in our midst, but as a social expression of values. Because of that, when we address the all-important issue of inequality, we cannot “bracket” out technology as an independent variable, as a machine with its own, bounded effects. Technology and the technological mindset permeate nearly everything we do these days. The ten guilty suspects on the “inequality express” that Bluestone identified are all aided and abetted by technology, just as if they were all carrying a gun or a using a bomb. To reform technology so that it helps turn back the trend toward greater inequality, we need to understand the social content of technological developments, and once we understand that, we can begin to reconfigure technology so that it serves everyone equally, instead of rewarding only the few.


  (1) Quoted in Harry Braverman, Labor and Monopoly Capital: The Degradation of Work in the Twentieth Century. New York: Monthly Review Press, 1974, p. 278.

  (2) Shoshana Zuboff, In the Age of the Smart Machine The Future of Work and Power. New York, Basic Books, 1988.

  (3) Jacques Attali, Millennium: Winners and Losers in the Coming World Order. New York: Random House, 1991, p. 101.

  (4) Barry Bluestone, “The Inequality Express,” The American Prospect, 20 (Winter 1995), pp. 81-93.

  (5) William Julius Wilson, When Work Disappears: The World of the New Urban Poor. New York: Knopf, 1996, p. xiii.

  (6) Bluestone, p. 89.

  (7) Robert H. Frank and Philip J. Cook, The Winner-Take-All Society. New York: Penguin, 1996.