The World As A Single Machine

Manufacturing Survey

Iain CARSON,

The World As A Single Machine

Rich countries worry that manufacturing is passing into the hands of poor countries. It is, but it’s not a worry, says Iain Carson CHANGE always frightens people. And today the world’s economy is going through two great changes, both bigger than an Asian financial crisis here or a European monetary union there.

The first change is that a lot of industrial production is moving from the United States, Western Europe and Japan to developing countries in Latin America, South-East Asia and Eastern Europe; the classic example is the exodus of textiles from the rich world over the past two decades (see 1). In 1950, the United States alone accounted for more than half of the world’s economic output. In 1990, its share was down to a quarter. Even then, North America, Europe and Japan were between them still producing three-quarters of the world’s output. But, as DeAnne Julius and Richard Brown, two economists working in Britain, pointed out in a famous essay in 1993, times are a-changing. Quite soon now, many big western companies will have more employees (and customers) in poor countries than in rich ones.

The second great change is that, in the rich countries of the OECD , the balance of economic activity is swinging from manufacturing to services (see 2). In the United States and Britain, the proportion of workers in manufacturing has shrunk since 1900 from around 40% to barely half that. Even in Germany and Japan, which rebuilt so many factories after 1945, manufacturing’s share of jobs is now below 30%. The effect of the shift is increased as manufacturing moves from rich countries to the developing ones, whose cheap labour gives them a sharp advantage in many of the repetitive tasks required by mass production.

These trends have caused an agonised debate about the "deindustrialisation of the West". When the oil-price rises of the 1970s brought inflation to Europe and America, many people feared a rapid decline in manufacturing as output shifted to developing countries with cheaper labour. By the mid-1980s, a lot of Americans had come to believe that their country’s industry was being "hollowed out" as its basic activities moved to low-wage areas in Mexico and Asia. A sudden cancer, it seemed, had gripped the entrails of American industry.

It was not like that. A change was happening, but it was not simply a destructive change, and it had already been happening for quite a long time. For years before the mid-1980s, the structure of American industry had been altering. The familiar picture of solid old companies like IBM , General Motors and Ford pulling together for the greater good of corporate America had long since turned brown at the edges. International competition had arrived ages ago.

Much of what an innocent American consumer might have thought of as "Made in America" was by 1985 the product of factories in many different parts of the world. By 1990, 40% of IBM ’s employees were non-Americans; Whirlpool, America’s leading supplier of domestic appliances, made most of its products in Mexico and Europe, having cut its American labour force by 10%; and General Electric was the biggest private-sector employer in Singapore, where American-owned factories employed over 100,000 Singaporeans to make electronic components to be shipped to the United States. By the early 1990s about a fifth of the total output of American firms was being produced by non-Americans outside America. As 3 shows, foreign investment in the past three decades has risen faster than trade and world output.

American industry had for years been changing, as the low-cost production of things like textiles, clothing, shoes, handbags, car seats and electrical wiring migrated to Mexico. Much that wore a Detroit car maker’s badge or a computer company’s brand was really the product of an elaborate international web of suppliers and assemblers. But it took a couple of decades for the politicians to realise what was going on. When they eventually didas the car makers of Detroit and the computer companies of California imported more and more of their components, from axles to microchipsthere arose desperate cries about "Japanese dominance" of high-tech industries and the "collapse" of much American production. And it was not only in America that people were falling into despair.

There is something curious about this. Why did French and Belgian politicians get themselves into a lather when Renault, France’s leading car maker, decided to close a factory near Brussels last spring? Why do fears of similar protests lead General Motors and Ford to keep surplus capacity across Europe, when a closer look at how many cars they can actually sell should lead them to axe their least efficient factories?

The broad answer, especially in Europe, is that many people still vaguely believe that manufacturing somehow matters more than any other economic activity; that making things you can drop on your foot is in some way superior to fingering a computer or cutting somebody’s hair. Never mind that more than two-thirds of output in the OECD countries, and up to four-fifths of employment, is now in the service sector. Making things in factories is still what real men do (as, 150 years ago, growing things in fields was what real men did). Quietly strip 3,000 jobs out of a national network of retail banks, and no one will raise an eyebrow. Open a telephone sales centre in the north-east of England, creating 3,000 new jobs, and it might get a mention on the local news. Close a steel mill, and the gasps of dismay go on for weeks. Open a semiconductor plant, and the hosannas echo for months.

For most of today’s rich countries, there was indeed a period when economic success was synonymous with manufacturing. Plenty of Britons who have no memory of the real thing still feel a pang of artificial nostalgia about Birmingham as "the workshop of the world" and Clyde-built ships that ruled the waves. Moreover, success in manufacturing was linked to geopolitical power. The democracies were able to defeat Germany in two world wars because America’s industrial machine poured out such a flood of tanks and warships and bombers. Best of all, in some ways, manufacturing was for long a source of reasonably reliable and well-paid jobs for young men with plenty of muscle and little else. It still is, to some extent. "We employ the guys who are never going to be Microsoft programmers," says one manufacturing boss at Chrysler.

It Isn't Special

So should the rich world worry that its manufacturing sector now seems to be migrating to low-wage competitors? Without big factories that ship steel, cars, machine tools and television sets to foreign customers, how can rich countries earn their keep in the world, finding the wherewithal to buy their food and oil and other raw materials? Ms Julius and Mr Brown, in that 1993 essay, called this the "Manufacturing is Something Special" argument. Manufacturing, in this way of looking at things, brings more growth, better-paid jobs, fatter export earnings and greater technological progress than any other economic activity.

Not so, the Julius-Brown essay explained. A household can use only so many cars and refrigerators and dishwashers in its members’ lifetime. As countries get richer, a rising share of income goes on holidays, health and education. Busy people want to hire other people to clean their homes, launder their clothes and so on. Anyway, many jobs traditionally thought of as part of "manufacturing", such as the design and marketing of products, are really service jobs. As demand for them increases, these service jobs become better paid and more interesting compared with the drudgery of factory workmuch of which is in any case moving overseas. Services are growing fast as a component of international trade, encouraged by widespread deregulation. The chunky, capital-intensive making of cars, chemicals and computers is starting to look middle-aged. If you want the vitality of youth, turn to things like telecommunications, aviation, biotechnology or the health-care industry.

In all rich countries, manufacturing’s share of total output is shrinking, and its payroll is shrinking even faster. The same thing happened with agriculture. At the beginning of this century, 68% of Japan’s labour force worked on the land, 44% of America’s and about 20% of Britain’s. Today, agriculture accounts for only 7% of workers in Japan, 3% in America and 2% in Britain. Yet the fading of agriculture did not bring impoverishment to these countries. Nor will the fading of factories. There will be other things for their people to do, which will bring comfortable incomes and may anyway be more interesting than hoeing fields or pulling levers on production lines.

This survey will look at the ways in which manufacturing is changing in both rich countries and poorer ones. Its various stages are becoming separate and dispersed, rather than being under one roof, or inside one company. The distinction between manufacturing and services is getting blurred. And, not least, new computer software is helping companies to organise themselves better around the task of serving their individual customers; this is rewriting the rules of mass production.

All this feeds back into the structure of manufacturing. Whole industries no longer migrate, as shipbuilding did from Europe to Asia in the 1970s. Drawing on examples from the industrial heartland of America, from the maquiladora factories in Mexico and from the Hong Kong-based manufacturing network, the survey will argue that today manufacturing is becoming a genuinely international affair. The fancy work gets done in rich countries by skilled workers, the simpler parts elsewhere in the global supply chain. More and more of the process is handled by multinational companies, quick to see what is best done where. There is nothing to fear in this. Any country that is willing to use the skills it possesses will gain from joining in.

The Way It Was

DATE 20-Jun-98 The Economist

Manufacturing was a cosy business when countries looked after their own. Then came competition AT THE start of the 20^th century, industry in America, Europe and Japan was beginning to assemble itself into large groups. America’s cartel-busting Sherman act of 1890 fell victim to the law of unintended consequences by promoting a merger boom. If competitors could no longer collude, they could combine; and they did.

Robert Reich pointed out in his 1991 book, "Work of Nations", that as early as 1904 a third of all American manufacturing was carried out by 318 firms with a combined capital of $7.3 billion ($121 billion in today’s terms). Tariff barriers kept up domestic prices, so it often made sense to buy up your supplier. When U.S. Steel found it expensive to import coal and iron ore from Canada, it bought iron and coal producers in Pennsylvania. Vertical integrationhuge corporations controlling every stage of production from raw materials to finished goodshad been born.

Add the moving assembly line, and you had cheap mass production. Henry Ford brought in assembly-line production in 1913 after visiting the Chicago slaughterhouses. He had watched cow carcasses being carried down a line on which they were systematically stripped and hacked to pieces. He decided he could make large amounts of money by doing it with cars instead of cows, and using the technique the other way roundstarting with small components, and building them into an entire car.

This was how General Motors and then U.S. Steel, General Electric, American Can and International Harvester were born. In Japan, similar consolidations produced zaibatsu giants like Mitsubishi and Kawasaki. In Europe came Siemens, Brown Boveri, Philips, Compagnie Generale d’Electricite, Imperial Chemical Industries.

These companies became national champions, and they had an unwritten contract with their governments. America’s big companies organised the production of goods in huge volumes; the economies of scale brought down the costs of production; the American purchaser could go shopping more cheaply. Prices would be set high enough, of course, to provide for investment in new machinery and to give shareholders comfortable dividends (and to pay salaries and wages good enough to keep managers and workers reasonably happy). In return, the government would ward off competition from abroad, and tolerate some price collusion at home. By the time of the New Deal, in the early 1930s, these giant corporations were national institutions. In the end what was good for General Motors was indeed, up to a point, good for America. Only 4% of cars bought in America were made abroad; the rest came from the Big Three car companies in Detroit.

A similar process took place in Europe and Japan. In each industry, two or three big companies became dominant. Wages were set for the whole industry, rather than company by company, which limited competition in costs. With little competition from abroad, prices could be co-ordinated. As Mr Reich put it: "The system contained its own internal logic. Big Business, Big Labour and the public at large would subsidise high-volume production in order to gain greater efficiencies of scale, which in turn would employ a growing middle class of Americans capable of buying the expanded output. It was a truly national bargain."

But it was doomed. These same American companies had gradually put down roots abroad. Soon they discovered that they could make goods in their foreign subsidiaries, ship them to America, and sell them at a bigger profit. And the foreigners, learning the lessons of American capitalism, were ready to compete. American consumers began to see that out of Japan were coming cheaper cameras and television sets. Container ships and better telecommunications shrank the world. A growing range of goods could be made where it was cheapest to make them, and then shipped anywhere in the world. Here, at last, was true competition.

The Lesson Written In Red Ink

In the way of these things, people were slow to see what was happening. The reality for long remained buried in the accounts of the big corporations. Richard Schonberger pulls it together in a new book ("World Class Manufacturing: The Next Decade", The Free Press, $30). He cuts through obfuscatory accounting to focus on one basic indicator of the health of a manufacturer. How many times in a year does a company turn over its stock of the raw materials and other things it converts into finished goods? The faster the turnover, the more efficient the company; and, other things being equal, the more money it should make.

As chart 4 suggests, manufacturers’ performance in this matter has gone through a V-shaped pattern since 1945. For a couple of decades, their stock turnover grew slower and slower. Then began the necessary reacceleration: first of all, as you might guess, in Japan. The Japanese began the introduction of "lean" manufacturing techniquesthe habit of carrying minimal stocks, having parts delivered direct to the assembly line "just in time", and making sure the quality was right from the word go. It took American manufacturers until about 1975 to see the point, when competition from sharpened-up Japanese companies really began to bite. IBM ’s downward slide, having started later than that of some other big companies, also went on longer, reaching its nadir in 1984, Mr Schonberger says: that is how a once-great company was brought to its knees by newcomers in the personal-consumer market, such as Compaq.

Mr Schonberger reckons it was the growing adoption of Japanese "lean methods" that caused the improvement discernible in America from the late 1970s onwards. He cites star performances by Ford Motor, John Deere (farm machinery), TRW (an electronics and engineering conglomerate), Eaton (axles) and PepsiCo. He also gives good marks to Cummins Engine, Caterpillar, Black & Decker, Motorola (microchips and electronics products), General Mills (food products) and Honeywell (industrial control equipment).

Looking back, he sees the 1970s as Japan’s decade. Having invented lean manufacturing, it could invade the markets of America and Europe with a formidable array of manufactures ranging from steel plate and ships to cars, calculators and televisions. But then the tide turned. The Americans had caught the point. From the mid-1980s to the mid-1990s was America’s decade. Just when American magazines were writing agonised stories about the "hollowing out" of American industry and their country’s loss of competitiveness, American industry was starting to prove their lamentations wrong.

The rest of the survey looks at how these pressures affect manufacturing in different parts of the world, and how companies respond to them. The best place to start is the American Mid-West, the one-time "rust-belt" where in the mid-1970s the car industry and its suppliers in the engineering industries were falling into decline. It took an invasion of new capitaland new ideasto rescue them.

Motoring Lessons

DATE 20-Jun-98 The Economist

The world’s biggest industry, having had one revolution, needs another THE tremor that ran through Detroit in early May, when Chrysler said it was merging with Germany’s Daimler-Benz, was the latest of a series of shocks to the heart of the world’s car industry. The making of cars is not just the biggest single chunk of manufacturing. It is also something other parts of industry keep a nervous eye on. What the car world does can often win zealous imitation.

In the 1970s, most car factories were chaotic places. Outside, thousands of unsold vehicles sat waiting to be ordered. Inside, bits and pieces of would-be cars littered the floor among the fag-ends. Beeping fork-lift trucks wended their way up and down the line. There was plenty of hard physical work being done by half of the workers, admiringly watched by the other half, who had time to stand and stare. Even ten years ago, car factories anywhere were a noisy mess. In the older Nissan factories around Tokyo Bay, body panels used to be welded together by manic robots in a flurry of sparks and a deafening whirl of metal and machinery. Enter a Russian car factory today and you will see the same scene. Or go round General Motors’ ageing main assembly plant in the suburbs of Sao Paulo, in Brazil, and you will need headphones to hear what your guide is telling you above the clamour of the vast presses stamping out body panels with 300-ton blows, the power of a jumbo jet taking off.

But visit a modern Nissan factory in America or in north-east England, and you will relish the calm. The same is true in non-Japanese factories. In Europe’s newest big one, at Melfi, between Naples and Bari in the south of Italy, there is hardly any noise. The whirling robots welding the panels sit in a semi-basement. Engines, gearboxes, axles swoop off a conveyor to be pinned quietly into place. There are few workers in sight, except at the end of the assembly line where wheels, windscreens, dashboards, seats and interior linings are stuffed into the freshly made shell as it becomes a Fiat car. A finished car rolls off the line every minute.

A Slice Of The Lean, Please

The difference between the noisy, confused old factory and the smooth-flowing world of the modern ones is the subject of what may be the most important book about manufacturing of the past 20 years. When James Womack, Daniel Jones and Daniel Roos wrote "The Machine that Changed the World" in 1990, they were not writing about the motor car itself. They were describing the Toyota Production System ( TPS ), which had created the most efficient car-production system in the world. Much of what they first described in that book is now part of everyday business language.

The authors used the phrase "lean production" to describe the essence of the Toyota system, first developed in the 1950s by a Toyota manager called Taiichi Ohno. With "just-in-time" delivery came the biggest influx of Japanese vocabulary into English since the importation of judo. Soon factories in America and Europe were using kanban controlstags to control the flow of partsand were practising kaizen, the process of continuously seeking small improvements to gain greater efficiency, better ergonomics and higher quality.

Messrs Womack, Jones and Roos compare lean manufacturing with the old system of craftwork, and with Henry Ford-style mass production. The craft producer used skilled workers and flexible tools to make exactly what his customer wantedone piece of, say, custom-made furniture at a time. It was good, but it was dear. Mass production made things cheaper. It used purpose-built machinery operated by workers each of whom did one particular thing. To make sure the expensive machinery was used as productively as possible, a continuous flow of materials had to be fed into it. Costs fell dramatically, but production runs were inflexible; to vary the product even slightly, the expensive machines had to be re-set, and that could take ages.

The aim of lean manufacturing is to combine the best of both craftwork and mass production. It sets out to use less of each input: less labour, less machinery, less space, less time in designing products. Mass production concentrates on reducing defects to a tolerable level. Lean production seeks to eliminate all defects; if something goes wrong, a worker can pull a cord, a tandon, and the whole line stops while the fault is identified and put right. An old car factory would have produced a complete afternoon’s worth of cars with the same defect. In a lean factory, the mistake is nipped in the bud.

It is also more economical. Lean manufacturing rejects the old idea of making things in huge batches, which requires holding large buffer stocks of materials and components between each stage of the production process. Now each act performed in the factory is, as it were, done on demand. When so many cars of such-and-such a model are ordered, the body panels, engines, gearboxes and wheels are sucked on to the production line just in time to keep pace with the assembly of the whole car. The process eliminates waste by making only as much as is wanted at any given time; gone are the costly piles of work-in-progress that used to litter the factory floor. The change has doubled productivity in some car plants in America and Europe.

Lean production came to America and Europe when Americans and Europeans observed Japanese methods at close hand in the factories that Honda, Toyota and Nissan had built to assemble cars in their countries. The first serious go at the idea outside Japan was a joint venture between Toyota and General Motors at Fremont, near San Francisco. It destroyed the widespread idea that the Toyota system could work only with squads of humble, drilled Japanese. The Fremont experiment showed that Americans could make cars as efficiently as Japaneseif they were shown how.

At first, American companies were slow to digest the lesson. But the Japanese had by now plenty of their own factories in America. They kept clear of Detroit: the closest they have gone to the traditional heart of the American industry is the original Honda factory in Marysville, Ohio, and since then Nissan and Toyota have headed farther south, to Tennessee and Kentucky, to sites beyond the reach of the United Auto Workers. Helped by the fact that only three of the 16 foreign-owned car factories in America have union representation, Japanese car production in America has soared from 1m a year in 1989 to 2.5m today.

America’s Big Three had no choice but to change their ways. Japanese cars made in America, unhindered by import duties or voluntary export limits, could compete head-on with Detroit. So America embraced the revolution in manufacturing techniques. Europe, as usual, followed America. Nowadays American car production, if still less efficient than Japan’s (see 5), is far more competent than it used to be; the same is true of Europe’s. And the art of lean manufacturing is spreading from the factory into other parts of the business, and from the car world to other industries. Once again, tomorrow is looking brighter than yesterday.

Caterpillar’s Comeback

DATE 20-Jun-98 The Economist

Caterpillar’s Comeback

WHAT do you do if your main competitor is a mighty Japanese company, if your main factories are in the part of America where the United Auto Workers’ writ runs, and if you have been losing $1m a day for three years? That was the question facing Caterpillar, a company that makes some famous earth-moving equipment, back in 1987. It had to do something, or die.

So Caterpillar’s bosses took on the union, and struggled through America’s longest strike, which started in 1991 and went sporadically on until the end of 1995. In March this year the company eventually got the flexible, seven-year deal it wanted with the UAW ; in return, wages and conditions have been improved and the strikers are being employed again. In April Caterpillar reported a pre-tax profit of $2.4 billion on sales of $19 billion, a rise in profits of 155% over five years. How did it do it?

Although nearly half its sales are outside Americait also has factories in Belgium, Brazil and Japanthe heartland of Caterpillar is Illinois, home to 12 of its 38 American factories. Twenty years ago you could not see across the welding hall of the plant in Aurora, Illinois, because of the smoke. Today, the welding hall is completely clear; the giant slabs of thick sheet steel are quietly cut into shape by high-voltage plasma guns, which produce a much more precise cut and no smoke.

It used to take 6,000 workers 25 days to get one big back-loader through the plant. Today 3,000 workers can get it through in six days. They are not supermen, just rangy mid-westerners sporting the grizzled Kris Kristofferson look and an air of unhurried effort. But the way they work is different.

No longer do piles of metal parts lie around. Inventories have been cut by 60%, and the floors are clear. Some parts are carried silently overhead on conveyor belts; heavier sub-assemblies are driven around in ghostly self-guided vehicles ( SGV s). These creatures are summoned up by an operator at a computer terminal who notes that a given chunk of metal has finished being drilled, milled, ground or bored and is ready to move on to another place, where one of 47 robots will weld it to another chunk. The SGV finds its way around with the help of laser beams and bar-codes at every junction on the production lines. Caterpillar started spending $250m on such equipment ten years ago to make Aurora a "plant with a future".

The giant things made in Aurora are the responsibility of Bob Weng, a product manager with worldwide responsibilities. Until Caterpillar was transformed, he would have been flanked by colleagues responsible for engineering, production, marketing and so on. Now he takes the decisions. The new set-up, he says laconically, "clearly connects actions and words."

In the old days, it could take up to ten years for Caterpillar to design and introduce a new machine. Now it takes as little as 27 months. Engineers go out to building sites around the world to stand and stare at customers using their products, to see if they can work out ways to improve them. The Caterpillar people do not find that phrases like "customer-based processes" trip lightly off their tongues. But the company that came back from the dead is a fine example of how to be a successful global contender in modern manufacturing.

No Factory Is An Island

DATE 20-Jun-98 The Economist

Multi-coloured widgets to gonow!, says the market MANUFACTURING used to be pretty simple. The factory manager or the production director rarely had to think about suppliers or customers. All he did was to make sure that his machinery was producing widgets at the maximum hourly rate. Once he had worked out how to stick to that "standard rate" of production, he could sit back and relax. Customer needs? Delivery times? Efficient purchasing? That was what the purchasing department and the sales department were there for. Piles of inventory lying around, both raw materials and finished goods? Not his problem.

Now it is. The 1980s were the decade of lean production and right-first-time quality management. In the 1990s the game has grown even tougher. Customers are more and more demanding. They increasingly want the basic product to be enhanced by some individual variation, or some special service. Companies sweat to keep up with their demands, in terms both of the actual products and of the way they are delivered. A world-wide survey of over 900 manufacturing companies by Deloitte Consulting has found that, though the companies’ products seemed better than ever before, customer dissatisfaction has kept on growing.

The solution lies in "mass customisation". This means making basically similar products in hundreds, even thousands, of variations to suit specific customers’ needs. Some customers want deliveries in small lots at short order. Others will take bigger deliveries less frequently. Either way, of course, they want it at mass-production low prices. The only way to have any chance of satisfying them is to spread lean-production techniques throughout the whole company, not just the factory.

The factory no longer stands aloof from the grubby business of haggling with suppliers and customers. "About 15 years ago," recalls Dale Marco, a senior consultant at A.T. Kearney, a Chicago firm that specialises in manufacturing practice, "we began to see you couldn’t really separate physical distribution and manufacturing. From there came the recognition that you had to see them both as part of a supply chain."

The new way of doing things connects the factory to its suppliers upstream and its customers downstream. Daniel Jones and James Womack, two of the authors of "The Machine that Changed the World", returned two years ago to lean techniques in a second book, called "Lean Thinking" (Simon & Schuster, £16.99 or $26), which applies the idea to the whole company. They examined, for instance, the various processes required to put a can of coke on the shelf of Tesco, a British supermarket group: the "value chain", as they called it.

A team of researchers from Cardiff Business School found that the chain of actions required to make the canstarting at the bauxite mine in Australia, and passing through the various smelting and rolling processes to the manufacture of the can itself, printing its label, filling it with the cola drink and getting it into somebody’s refrigeratortook no less than 319 days. Making the can took far longer than making the coke. But, even so, only three hours of that time was spent in doing something that actually added value. The rest was spent in storage and transport; as many as 14 storage lots and warehouses were involved.

The blame for this apparently wasteful way of doing things lies with the huge and inflexible machinery needed for long production runs, in order to capture the economies of scale. This is the so-called "batch and queue" system common to most forms of manufacturing. Unfortunately, it requires the piling up of large buffer stocks of raw materials and component parts to make sure the machinery can be continuously fed. Each of the seven main links in the cola-bottling chain naturally seeks to maximise its own return on capital.

The aim of the Tesco analysis was to see whether some of these economies of scale, at one or another of those seven links, was outweighed by the waste involved in all the handling and storing required. Since the book came out, Mr Jones says that Tesco has made some progress in reorganising its shipments so as to cut out wasted time. "The way to do this", he says, "is not to start at the raw-material end but to work your way up the chain from the store." In time, he suggests, some of the suppliers may use smaller and cheaper machines, in themselves less economical than mass-production ones, which may nevertheless cut the cost of the total operation.

The Magic Offered By Software

As companies tear down the walls between the different parts of their work, they are realising that the various parts need to share the same flow of information. Hence the explosion in manufacturing software programs which let them integrate their financial data with payrolls, manufacturing and inventory records, purchasing information and the rest of it. This is known as Enterprise Resource Planning, ERP : the magic tool-kit needed to spread lean thinking throughout a company. Would-be lean companies, from Toyota to the latest recruit, are scrambling to have ERP s installed on their personal-computer systems. The software comes from companies such as SAP , the German leader in the field, Baan, PeopleSoft and Oracle.

An ERP is supposed to take into account all sorts of numbers, so that a company can know how efficiently it is using its various resourcespeople, money, machinesto satisfy its customers. If all aspects of the company are recorded in the same software, it is easier to keep the whole manufacturing operation in balance and to keep work flowing smoothly through the factory. Bottlenecks and imbalances show up quickly, and can be put right.

Needless to say, for every dollar a company spends with a software company to buy its ERP program, it finds itself paying another six to a management consultancy to adapt the program for its factories. No wonder these consultancies speak of ERP in reverent tones. But putting all the company’s information into one giant software program is not without risks. "It’s like pouring concrete over your business plan," says one now-sceptical expert, David Upton of Harvard Business School.

ERP s may or may not be the answer for companies that want to become leaner and nimbler. But there are other means of improving the way they buy their components and their raw materials. The Internet is giving a huge boost to the process. By using the Internet, or various private extranet systems based on it, companies can deal more directly with their suppliers to improve deliveries, stock levels, designs and lead times. Even better, they can use the networks to get lower prices by holding electronic auctions for the supply of basic components.

Leading the way is a system known as Automotive Network Exchange, which America’s Big Three car companies started as a pilot scheme in January and are due to put into operation with most of their suppliers this summer. Next year they plan to extend it to Europe. Any company wishing to sell to the Big Three will then have to get online, and be ready to settle down to some serious electronic bargaining. Other industries have happily taken up the idea. Boeing already runs its spares business through an extranet. General Electric has one of the biggest systems in the whole of manufacturing for dealing with its suppliers.

All this should produce huge benefits. It will make it easier and swifter for the assembly-line companies and their suppliers to work in partnership. Engineering designs on the web will save time and money by cutting out a whole series of meetings and consultations. By turning many car parts into commodities that can be bought and sold at auction, the web will make the market more open, competition keener and prices lower. The old ways of buying and selling were time-consuming and expensive. The new way opens the door to the "mass customisation" that lean producers now dream of.

None of this will make a byte’s worth of difference, of course, unless manufacturers are nimble enough to seize the opportunities technology is offering them. After years of downsizing, rightsizing and all the rest of it, this might seem just another pointless redrawing of organisational charts. Not at all, Mr Schonberger vigorously replies.

He maintains that the essential thing is for everybody in a company to understand how a business runs itself successfully. Out, he says, with management by edict and by procedures: that is just a way for bosses to tell everyone below them what to do. In the new century, Mr Schonberger wants manufacturing to be a process in which the only objective is to meet the customers’ needs, in which all the assorted kinds of workers feel involved in the task, and in which they are able to make use of an array of facts and data supplied on electronic platters.

So how does a company get this way? By organising production around customers, not the other way round. Sometimes this can mean moving halfway across the world. Not only do companies have to chase cheap labour for the simpler parts of their new global supply chains; they also have to bow to the customers’ demand that they provide product and service on the customers’ doorsteps. The next two articles look at what this is likely to mean in the next stage of the manufacturing story.

Bliss It Ain’t

DATE 20-Jun-98 The Economist

ENTHUSIASTS for lean production in factories often make it sound as if it were heaven on earth for workers as well as managers. For Daniel Jones and James Womack, the authors of "Lean Thinking", lean production is all about jolly teams of problem-solvers working together to improve products and processes. If only. They should watch the stampede for the door when the hooter sounds for the end of a shift.

Richard Delbridge of the University of Cardiff Business School has made a systematic study of the reality on the shopfloor, and tells the story in a recent book, "Life on the Line in Contemporary Manufacturing" (Oxford University Press, £35). A sociologist, Mr Delbridge spent several weeks working in two factories, one a car-parts maker in South Wales, the other a television assembly plant in the south of England.

Both, more or less, used Japanese lean-manufacturing techniques, just-in-time production, total-quality management and so on. But the shopfloor workers were less than fully committed to the view that their interests and the companies’ were one and the same. Mr Delbridge notes their chafing at the endless attempts to shave a fifth of a second off some task on the line, and their dislike of being bossed around. "They treat us like children the way they talk to us," explained a mother of four at the television factory. Another, after saying that they were supposed to work as a team, complained of the managers: "They should try doing these jobs. They just stand there over you, telling you what to do. They just don’t understand it."

The workers in the car-parts company seemed to have kept more control over their life on the line. But that did not stop resentments building up. Just before Christmas, a quality inspector thought of as a management lackey was tied up and dumped in a rubbish bin. He was rescued by a watchful security man. But one of his colleagues was put into a consignment of parts headed for Birmingham, and found only at a gate inspection. Just in time, indeed.

The Tijuana Triangle

DATE 20-Jun-98 The Economist

Mexico’s northern border is modern manufacturing on the move ONE night in 1959, a Mexican teenager was at a party in San Diego in southern California. Among the cool new American friends 19-year-old Enrique Mier y Teran made that night was a 20-something American who said he was in business. What business? "I have a factory making pincurl clips," said the American. This was still the 1950s, and the wannabe Rita Hayworths of those days needed something to control their huge hair bangs. Enrique laughed out loud. Offended, his new friend invited him to visit his factory the next day.

As he went round the plant, he asked the American how much he paid the 500 girls who operated the simple machinery and assembled the plastic clips. The answer was $65 a week. "I could get people to do that for 16 Mexican dollars a week," said Enriquea 50^th of what the American was paying. So the American lent Enrique two machines, and he started making pincurl clips in an old shed just over the border in Tijuana, to sell in America. He paid no Mexican duties on the imported parts, because he was in a free-trade zone by the border; and the American customs were persuaded that the plastic parts were going south for "repair", which got round American duties on re-entry. As the business grew, he got his businessman father to help him. Thus was born one of the first maquiladora factories along Mexico’s border with the United States. Now Mr Mier y Teran owns an industrial park and a consultancy which advises multinationals around the world how to do today what he started doing 39 years ago.

The locals say this is the busiest border in the world. Last year 56m people and nearly 1m laden trucks crossed into the San Diego area from Mexico (they don’t count the southbound traffic). A huge sprawl of maquiladora factories have grown out of those pincurl clips. Today nearly 4,000 such firms employ nearly 1m workers in Mexico, making goods whose value-added worth is over $7 billion a year. That puts them second only to oil in the Mexican economy. Almost half of them make textiles or consumer electronics. They are here because this border is a place where the rich world and the poor world touch hands. With Mexican wages only a quarter of American ones, that permits a spectacular expression of the law of comparative advantage.

Put it another way. In deciding where to put a new factory, a company has two main things in mind. It wants the cheapest possible labour costs. But it also wants to be close enough to its markets to respond to them swiftly and not to spend too much on getting its goods to them. The Mexican borderland provides American companies with a wonderful way of combining those two desirable things.

Matters have changed since 1959, of course. In the pincurl-clip days, when production was very basic, American companies would ship part-finished textiles or toys across the border so that hundreds of little Mexican sweat-shops could do the labour-intensive parts of the manufacturing job and then ship them back, sometimes before sundown the same day. The business grew more sophisticated when it was agreed that the materials and parts needed for manufacture could be imported free of normal duties for up to a year. When the finished goods produced by Mexican labour are exported from Mexico, duty is now charged only on the value added in that country. Bingo. Cheap labour, markets to hand, and a relaxed approach to what you are obliged to pay the government.

Among the early arrivals were American car-parts makers seeking cheap and nimble Mexican fingers to assemble seat covers, or tie together the kilometres of electrical wiring that go into every modern car. Then came American car-assembly plants. And in the past few years the Americans have been followed by a wave of investment from Japan and other Asian countries. These came for the same two reasons that had pulled in the Americanscheap labour and modest transport costs. The Japanese found they needed to pay local workers little or no more than they had been paying those in South-East Asia. The South Koreans joined in when their wage costs at home started to climb. They both relished the idea of producing things inexpensively on the doorstep of the United States. So now you have the Tijuana triangle: American market, Asian investment, and Tijuana production.

The area around Tijuana, in the Mexican state of Baja California, is home to the biggest group of maquiladoras, about 1,000 firms employing 200,000 workers. Across the low, bare hills on the eastern outskirts of Tijuana spreads a rash of new factories and housing developments. The big products here are consumer electronics, notably television sets, of which some 15m are exported every year, mostly to the United States.

Among the Korean companies operating in the area is Samsung, here since 1988, which makes cameras and television monitors in a huge $2 billion plant. Hyundai has a large factory that manufactures transport containers and lorry trailers. The watch-towers manned by guards which ring these huge complexes give them an uncomfortable resemblance to labour camps. The Koreans’ management style can be harsh; protests against it have led to a number of fierce labour disputes.

Japan is represented by consumer electronics firms such as Sony, Hitachi, Matsushita and Sanyo; the latter is the area’s biggest employer, with 4,500 workers. About 18 months ago Sanyo North America moved its headquarters from New York to the outskirts of San Diego. There sits its chairman, Motoharu Iue, just round the corner from Sanyo’s little factory on the American side of the border, where 100 workers produce a daily output of 800 mini-fridges (the sort that distinguish some hotel rooms from prison cells). A mile or so down the road 650 Mexicans produce 6,000 of the same things a day, for wages that are barely a fifth of what the workers in the American factory get. That the latter still exists seems rather quixotic.

Sanyo’s biggest product in Tijuana is television sets; its output of 3.5m sets a year is beaten only by Sony’s 6m. Mr Iue says that his company used to do all its television work inside the United States; now it does only some final assembly in Arkansas. Like many bosses in Tijuana, he is keen to tell you that wages are a small proportion of his total production costs, around 10%, and that the real attraction of Tijuana is cheap materials and components and the fact that it is easy to run just-in-time production systems here. This is somewhat disingenuous. One reason why wages are so small a part of total production costs is that the wages are so low. And his suppliers can provide those cheap components partly because they too pay such low wages.

But Evening Approaches

As you walk around Sanyo’s television plant, though, you are struck by the amount of automation used in making the "mother boards" that are the electronic innards of a set. Row upon row of state-of-the-art automatic insertion machines whirl and turn their little working heads to pop 16,000 pieces of circuits and micro-electronic devices into place every hour. Downstairs, where some of the clunkier gadgets (things the size of an aspirin tablet) are plugged in by hand, there is another sign of creeping automation. One line has about 20 workers; but another has machines doing the jobs formerly done by 18 employees, with only one worker stationed at each end to make sure all is going well.

Which shows that even in places like Tijuana, the heartland of the maquiladoras, the stern laws of economics are at work. Once purely an area of low-wage, low-value-added production, it is now moving to a more developed form of industrialisation.

Unemployment is down to just over 1% (compared with an official 4.5% for Mexico as a whole). Labour turnover exceeds 15%, as workers hop from one factory to another looking for a better deal. Economists at the University of San Diego say that more than 12% of the workers in Baja California are now classified as technicians. Sanyo’s Mexican managers say their growing automation requires a strong corps of experts to set up and maintain the expensive equipment. Wage competition between companies still mainly takes the form of fringe benefits, but the total cost of labour is creeping up. The simple, labour-intensive work the Tijuana area started with (sorting soap-powder coupons for Procter & Gamble, doing wiring looms for General Motors) has long since drifted off to the poorer parts of Mexico, or to other cheap-wage Latin American countries such ase Guatemala.

The pressure is growing. The rules of the North American Free-Trade Agreement ( NAFTA ), which embraces the United States, Mexico and Canada, will soon require many of these factories to buy 60-80% of their raw materials and parts inside NAFTA ’s borders, or face import duties. Since the United States and Canada now produce virtually nothing in the way of consumer electronics, this will increase the cost of the television tubes and other components that companies like Sanyo buy from Asia, unless that NAFTA rule gets changed (and unless the devaluation of East Asian currencies caused by Asia’s recent economic troubles cancels out the cost of the new duties). In the cut-throat market for television sets, that could make all the difference.

Moreover, those Asian devaluations make Asia a sharper competitor for the maquiladoras; their low-wage advantage is now under challenge. Mr Iue has been doing a study of this for his colleagues back in Japan. He is disinclined to shift production back to Asia, mainly because that would hugely lengthen the distance between his factories and their market in America. "But it is a big problem," he is obliged to admit.

Pretty clearly, things will be different in Tijuana five or ten years from now. The existence of NAFTA , a free-trade area with 390m people and $8.8 trillion-worth of output (twice as much as Japan’s), will slowly make it a much more sophisticated economy. Mexico’s borderland will no longer be just a home for the maquiladoras. But history will be grateful for the lesson the maquiladoras have provided. So will be the other parts of the world to which their successors have profitably migrated.

The Ever-Spreading Tentacles Of Hong Kong

DATE 20-Jun-98 The Economist

What the world’s third-richest place has to teach YOU would have thought that being one of the richest places in the world, a hub of international trade that sits at the gateway to China, would produce a certain satisfaction among the 6.5m inhabitants of Hong Kong. But even in this apparent economic paradise people worry, and argue, about where paradise goes next.

In one corner of the argument, fighting for the service-economy idea, are the team of authors who last year wrote "The Hong Kong Advantage" (Oxford University Press, £20). They say that this city-state’s prosperity will continue if it concentrates on being a centre of financial and business services and a provider of culture and entertainment for the Asia-Pacific region. The Harvard professor who led the team, Michael Enright, has voted with his feet and moved to Hong Kong University. In the other corner, the authors of "Made by Hong Kong" (also Oxford University Press, £20), who come from the Massachusetts Institute of Technology, retort that Hong Kong will stay successful only if it holds on to its manufacturing prowess and takes manufacturing up-market into high-tech products.

It was manufacturing that made Hong Kong’s fortune. When more than 1m refugees from the Communists’ victory in China’s civil war flooded over the border in 1949, the British colonial authorities had to do something to keep them alive and Hong Kong afloat. Cheap land for building factories, and public housing, were the answer. Helped by the talents of entrepreneurs who had fled from the mainland, the rescue operation worked. By the early 1950s Hong Kong was supplying the world with T -shirts, plastic flowers and flip-flop sandals. The big industry was textiles, but handbags, toys, watches and shoes also poured out of tiny sweat-shops for export to America and Europe. By 1961, nearly half the workforce was employed in manufacturing, making a quarter of Hong Kong’s GDP .

But China’s Communists halted the flow of refugees, and Hong Kong started to run short of labour. It was then that the Chinese government took a historic decision which changed the way manufacturing operates in this part of the world.

In 1978, China opened up some of its coastal regions, including the area on Hong Kong’s border, to foreign investors. Suddenly, Hong Kong’s businessmen had a new supply of cheap labour and new land for factory-building. Most of the light manufacturing that used to take place in Hong Kong crossed the border into China. By 1997 manufacturing’s share of Hong Kong’s jobs was down from nearly half to a mere 15%; its share of output fell to a mere 8.8% (see 6). But over the border, around the Pearl River delta, 5m Chinese are now employed in factories owned by Hong Kong businesses. Today Hong Kong has $72 billion invested in China, nearly all of it in manufacturing.

Two examples show what has been achieved by this leap across the border. Allen Wong runs VT ech, a Hong Kong company that has 60% of the world market for kiddies’ computers and other electronic learning devices, and also makes cordless telephones. Its sales have grown from an annual $150m ten years ago to over $800m today. In 1987, when he found he could not get enough workers or cheap land in Hong Kong, Mr Wong moved production over the border into Guangdong province.

He now employs 22,000 Chinese women in two factories, most of them brought in by bus from Sichuan province deep inside China (because the wages the locals want are already too high). The women get 500 yuan ($60) a month, plus board and lodging, compared with the HK$6,000 a month ($770) people get for similar work in Hong Kong. Now Mr Wong is scouting for even cheaper labour in Thailand. Hong Kong is the conduit through which his exports pour. Distance from the eventual market does not matter much to him. Because his products are mostly bought by parents for their children, demand is pretty stable; he can predict sales, he says, for about two years ahead.

But not all Hong Kong-based businesses can afford to be so relaxed about distance from the market. Another company that crossed to the mainland is Johnson Electric, which makes the tiny motors that go into electric toothbrushes, CD players and many other things. The average Mercedes car has 80 such micromotors to open its sun-roof, adjust its mirrors and move its seats.

Under Patrick Wang, whose father founded the business 39 years ago, Johnson has tripled in size in the past ten years. It now has sales of HK$2.6 billion ($33om) in 20 countries. It employs 13,000 people, most of them in Johnson Electric City, two huge factories 50 miles from Hong Kong in the Pearl River delta with their own dormitories, canteens and sports centres. The peasant girls who work in Electric City earn around 500-600 yuan a month, plus food and board: again, only a fraction of what a similar worker would get in Hong Kong. Yet they save much of what they earn. After a few years in the factory, they can go home to their villages with a dowry to help them look for a husband.

In Hong Kong itself Johnson employs only 200 workers, mostly engineers designing new products. They work in teams that have to keep in close touch with the customers. A car manufacturer, for instance, will want Johnson to come up with a little motor ideal for the particular job it has to do in a given car. Mr Wang says that ten years ago customers asking for design changes were happy with a two-week response by telex. Today, they want an e-mail within two hours.

This has made Mr Wang change his organisation. The problem of distance looms larger every day. "Two weeks in a container on the sea was bad enough in the days of telex response times," he says. "It is no use now." So he is investing in a new factoryin Mexico, a maquiladora on the doorstep of his biggest market, the United States. There he will be paying wages of around $2 an hour, about three times what he pays in China. But that is the price of being close enough to his market to meet sudden changes in what the customer wants.

Wider Still And Wider

In short, much of what began in Hong Kong and then moved over the border into mainland China is now having to think of migrating even farther afield. This applies in particular to the textile and toy industries. David Dodwell, one of the authors of "The Hong Kong Advantage", estimates that two-thirds of Hong Kong’s 30,000 producers of clothes and soft toys operate in two locations (usually China as well as Hong Kong) but that as many as a fifth now have plants in two places outside Hong Kong. The continuing search for cheap labour is not the only reason for this diaspora.

There are two disadvantages in selling things made in China through Hong Kong. For textile firms, trade continues to be regulated by the hoary old Multifibre Arrangement, which still contrives to protect the textile industry in rich countries where labour costs far more than in China. So China gets no quota for exports to Europe and America. And anybody exporting anything from China has to live with the risk that one day Congress in Washington will take away China’s most-favoured-nation trade status, which allows it easy access to the American market. So some manufacturers have started to move on from China, to put at least some of their production into countries where they can hope to disguise its Chinese origins.

Victor Fung is chairman both of the Hong Kong Trade Development Council and of the family firm of Li & Fung, a fine example of this new diaspora. Li & Fung is a spider’s web of manufacturing in 23 countries, with operations not only in various parts of South-East Asia but also in Latin America, Eastern Europe, the Caribbean and Mauritius.

Mr Fung explains how the system works. A foreign company will come to him with a modest producta ballpoint pen, for example, or a simple dressand ask him to find out where it can be made more cheaply than anywhere the inquiring company yet knows about. Mr Fung’s people set out to find not only a source of ever-cheaper labour but also somewhere safe from trade restrictions on Chinese production. Take that simple dress. The yarn may be spun in Korea, the fabric woven in Taiwan, the zips bought from Japan, and the garment part-finished in China before it passes through a final stitching-factory in Indonesia. "What we are doing is finding the best place for every operation," Mr Fung says. "At the same time, we are lining up factors of production so that we can cut lead times from three months to five weeks."

Li & Fung has a network of 7,500 regular suppliers, employing an average of 200 workers apiece. In other words, about 11/2m workers to some extent depend on this firm. Mr Fung uses only about a third of his network at any one time; but the system gives himand his customersa splendid degree of flexibility. The hub of the system is his team in Hong Kong, what he calls his "smokeless factory". The clothes business has always been keen on fast responses to changing tastes. Mr Fung reckons the habit is now spreading to other industries, even those making things like personal computers. This needs an international network which can make sure that all the parts needed for the final product are flowing smoothly, while keeping its inventories economically small. Mr Fung’s company has been doing this sort of thing for years"but now we see them calling it supply-chain management."

All these companies keep in Hong Kong the central functions of product development and engineeringthe front end, as it were, of the manufacturing supply chain. The back end, marketing and distribution, also stays in Hong Kong. Although these are officially classified as "service activities", they are really part of manufacturing in the modern, lean-all-through sense of the word.

Hana Technologies, part of the Thai Hana Group, which bought it from the Swire conglomerate a few years ago, still uses Hong Kong for some manufacturing in the old sense of the word. Its factory puts microchips into plastic packaging and fits them with terminals to link up with other parts of the end-product, a computer or whatever. It sounds simple, but this is high-precision work that needs sophisticated machinery, and often involves novel operations. Once any given operation is running smoothly, the company moves production to Thailand or China. Here is another example of the way Hong Kong is moving up-market. It does the difficult jobs of design and early production, and then hands the routine stuff over to others.

Taiwan Can Do It Too

What has happened in Hong Kong was not going to escape other people’s notice. The Taiwanese have caught on well. Back in the 1950s, Taiwan was part of the region’s plastic-flowers and flip-flop-sandals economy. Things moved on when state planning created a series of industrial parks along the island’s coast, each specialising in its own thingpetrochemicals and plastics (from imported raw materials), cement, steel and so on. But the economy really started coming alive when the government gradually stopped poking so many fingers into it. The bureaucrats no longer try to pick winners, though they still provide cheap science parks and some handy tax breaks.

Hence the blossoming of companies like Acer, a computer company which has its main factory in the Hsinchu science park, two hours down the road from the capital, Taipei. Acer is now the world’s third-biggest maker of personal computers. Its boss is Stan Shih, who started his business life selling ducks’ eggs off a street stall. Until 1993, all Acer’s computer mother boards, the basic circuitry, used to be assembled in Taiwan. But then, with local wages approaching $600 a month, and local labour so short that Acer had to ship in hundreds of Filipinos, it opened factories in Zhongshan in China’s Guangdong province and in Subic Bay in the Philippines. Now it has added another, in Mexico, and has assembly plants in Malaysia, Singapore, Indonesia, Dubai and Europe.

Productivity in the Philippine plant is only half of that in the Hsinchu science park, in the Chinese plant only a third. But then the wages the company has to pay in the Philippines are less than a third of those in Taiwan, and in China only a tenth. At the other end of the scale, Acer now has assembly plants in Britain, France and Italy. These are expensive to run but they are physically close to the market, so that the final product can be adjusted to meet local requirements. Mr Shih has got the message about distributed manufacturing networks.

The Hsinchu factory in Taiwan will not be extended; the company’s expansion will all happen abroad. Already, two-thirds of its output of mother boards and nine-tenths of its computers are made outside Taiwan. The Hsinchu plant, like those Hong Kong factories, is changing its role to concentrate on research and product development, plus the pilot production of new items. Maybe some of its more expensive products will continue to be made in Taiwan but most others, once production is proceeding smoothly, will go abroad.

It took Acer nearly 20 years after its foundation in 1976 to move to offshore production. Other companies are doing it more swiftly. About ten years ago Kuo Feng Corporation and Shamrock Technology, among others, started making personal-computer monitors in Taiwan. Soon Taiwan had getting on for 60% of the booming world market for these things, worth $15 billion a year. Now, as demand for them levels off and profit margins fade, their manufacture is seen in Taiwan as a sunset industry. But the sun is not vanishing into the sea; it is just moving over the water into China, where costs are so much lower. Both Kuo Feng and Shamrock plan to shift the bulk of their production to the mainland in the next 12 months.

Within a single decade, Taiwan has become the world leader in this field, and then almost abandoned it. Scotland’s Clydeside dominated world shipbuilding for three generations and then took a quarter of a century to hand the baton over to South Korea and Japan. The globe spins faster these days.

Post-Industrial Manufacturing

DATE 20-Jun-98 The Economist

Goodbye to the old dividing line between manufacturing and services THE old definitions of manufacturing are no longer worth much. One dictionary calls it "the making of an article by physical labour or machinery". The 1987 Standard Industrial Classification manual in the United States says a factory is "an establishment engaged in the mechanical or chemical transformation of materials into new products".

Now go and stand in a queue at a McDonald’s takeaway. Observe what goes on behind the counter. Grills cook raw discs of minced beef. Some workers are tending the grills; others are loading potato chips into vats of hot fat; yet others are taking orders, packing the output of their colleagues into cardboard cartons, adding whatever extras the customer calls for. Would you call this a service activityor the distributed manufacture of cooked-meat products? The hectic repetition of different tasks, the loading and unloading of the cooking machines, recall nothing so much as a car assembly line. If Charlie Chaplin were alive today, he would film "Modern Times" in a fast-food joint.

Alex Trotman, the chairman of Ford, sent a taskforce to McDonald’s when he set out to transform the car company four years ago, to learn how McDonald’s turns out the same burgers all around the world. Ford wanted to do the same sort of thing. It wanted to change itself from a collection of regional companies into one which designed and produced cars on a global basis, with a global supply chain a Hong Kong textile boss would recognise.

Mr Trotman has long believed that in future a car company will beat its rivals not so much by the shape of the car it sells, or the surge of its engine, as by what the company can do for its customers while they own the car. The people who make Mercedes cars agree. This summer’s new product from Mercedes (made in eastern France), the two-seater Smart town car, brings with it as part of the package several weeks’ rental per year of a larger vehicle.

The Industrial Revolution Comes Home

It is not just cars. IBM ’s recovery is largely due to the fact that it has become a service company as well as a computer maker. And mass-customisation, as the jargon has it, applies to lots of other things. Look at denim jeans. Years ago, the textiles industry shifted the production of this originally simple item of clothing away from America and Europe to places where the workers accepted lower wages. But nowadays there are jeans and jeans. You can get jeans made to measure after a body-scan takes your shape and a satellite sends the details to the factory. The new, quick-response market requires factories closer to the customer. So textiles manufacturing is inching back into America and Europe.

Or take made-to-measure shoes. In the old days, the price of such shoes reflected the high cost of handcraft work. But now a combination of new technology, lean production and supply-chain management is bringing in mass-customised shoesto the rescue of, among others, Italy’s shoe industry. Even a machine-tool maker can today make as much money out of installing his product, showing the buyer how to run it and then servicing it as he makes from the price-tag on the tool itself. In many manufactured goods, the cost of physically transforming the raw materials into the object sold may be only a tenth of the retail price.

This is good news for Americans and Europeans, the inhabitants of the rich world. For years, the making of many of the parts that go into motor cars, the cloth that goes into jeans, the circuit boards that go into computers, has been moving away from their countries into lower-wage regions. Multinational companies have fitted themselves out with the technologies and the management techniques needed to produce even pretty complicated things efficiently in places where the labour is relatively unskilled. When labour in Tijuana and Guangdong gets too expensive, production moves deeper into Mexico or China, or into India. If one day labour even there gets too dear, more and more manufacturing will have to be automated.

But it is not just geographically that manufacturing is moving on. The nature of manufacturingthe meaning of the wordis changing. Rich consumers in rich countries want things now, and they want them made precisely the way they wish. They are not content with mass-production uniformity. They will not always wait for a container ship that takes two weeks to cross the Pacific. This is giving birth to a new kind of manufacturing: a long chain of processes that can go right around the world; a global conveyor belt, as it were.

There are sound reasons why the higher-value parts of these new global chains will remain in America and Europe. In an era of mass customisation, the smooth and relaxed relationship that is possible between manufacturers, customers and suppliers in the electronic world of Silicon Valleyor the textiles and footwear world of Emilia Romagna, or the car-making worlds of Baden-Wurttemberg and Bavariagives a special new advantage to those who operate in such places.

Twenty years ago the economists and politicians of the Atlantic world talked of sunset and sunrise industries: shipbuilding bad, electronics good. Then they noticed that even some of the sunrise industries were migrating to places where wages were lower. Doom, it seemed, was closing in.

What they failed to notice until recently was the emergence of a global structure for manufacturing, whatever the industry. The law of supply and demand will always drive the lowest-value-added part of the supply chain to where wages are lowest. The tennis courts of Hong Kong are full each Saturday of perspiring Americans and Europeans taking a break from the endless search for cheaper Chinese factories to make widgets designed in Cologne or Chicago for the multinationals that employ them. Does this not mean, the gloomsters demand, that manufacturing in Europe and America is bound to shrink away to almost nothing, as farming already has done in so many countries?

Only in the narrow, outdated meaning of the word. In the new meaning introduced by the new technologies, much of what used to be called "service work" is becoming inextricably part of the same thing as what used to be called "manufacturing". It is the virtual part, as it were, of the total business. And it will bring solid rewards to those who practise itas Americans and Europeans, above all, are able to do. Manufacturers of the world unite. You have nothing to lose but your supply chains.