Supply chains

THE macroeconomic discussions that Apple's success prompts tend to be very curious things. Here we have a company that's been phenomenally successful, making products people love and directly creating nearly 50,000 American jobs in doing so, criticised for not locating its manufacturing operations in America, even as Americans complain to Apple about the working conditions of those doing the manufacture abroad: life in dormitories, 12-hour shifts 6 days a week, and low pay. It isn't enough for Apple to have changed the world with its innovative consumer electronics. It must also rebuild American manufacturing, and not just any manufacturing: the manufacturing of decades ago when reasonable hours and high wages were the norm.

The utility of Apple, however, is that it does provide a framework within which we can discuss the significant changes that have occurred across the global economy in recent decades. Contributing to that effort is a very nice and much talked about piece from the New York Times, which asks simply why it is that Apple's manufacturing is located in Asia.

You should read that piece for yourselves, but the story, in a nutshell, is this. Apple's products are assembed in massive factory complexes in China, run by Foxxconn, which also handles the production of consumer electronics for many other large players in the industry:

The facility has 230,000 employees, many working six days a week, often spending up to 12 hours a day at the plant. Over a quarter of Foxconn’s work force lives in company barracks and many workers earn less than $17 a day. When one Apple executive arrived during a shift change, his car was stuck in a river of employees streaming past. “The scale is unimaginable,” he said.

Foxconn employs nearly 300 guards to direct foot traffic so workers are not crushed in doorway bottlenecks. The facility’s central kitchen cooks an average of three tons of pork and 13 tons of rice a day. While factories are spotless, the air inside nearby teahouses is hazy with the smoke and stench of cigarettes.

Foxconn Technology has dozens of facilities in Asia and Eastern Europe, and in Mexico and Brazil, and it assembles an estimated 40 percent of the world’s consumer electronics for customers like Amazon, Dell, Hewlett-Packard, Motorola, Nintendo, Nokia, Samsung and Sony.

“They could hire 3,000 people overnight,” said Jennifer Rigoni, who was Apple’s worldwide supply demand manager until 2010, but declined to discuss specifics of her work. “What U.S. plant can find 3,000 people overnight and convince them to live in dorms?”

The components that go into the phone are quite often assembled in China, or elsewhere in Asia, as well:

Manufacturing glass for the iPhone revived a Corning factory in Kentucky, and today, much of the glass in iPhones is still made there. After the iPhone became a success, Corning received a flood of orders from other companies hoping to imitate Apple’s designs. Its strengthened glass sales have grown to more than $700 million a year, and it has hired or continued employing about 1,000 Americans to support the emerging market.

But as that market has expanded, the bulk of Corning’s strengthened glass manufacturing has occurred at plants in Japan and Taiwan.

“Our customers are in Taiwan, Korea, Japan and China,” said James B. Flaws, Corning’s vice chairman and chief financial officer. “We could make the glass here, and then ship it by boat, but that takes 35 days. Or, we could ship it by air, but that’s 10 times as expensive. So we build our glass factories next door to assembly factories, and those are overseas.”

All told, the physical production of Apple's products accounts for hundreds of thousands of manufacturing jobs. America, which finds itself several million jobs short of where it would like to be, and particularly short of the middle-skill manufacturing positions that once powered growth in the middle class, seems to want some of those back. Is that a reasonable desire?

How did we get here? A new paper by Richard Baldwin helps lay out some of the story. We can summarise it briefly by noting that specialisation is limited by the extent of the market. When transport costs are very high, the accessible market is quite small, perhaps no more than a village, which will then make nearly everything it needs for itself. As transport costs decline, economic activities may begin to unbundle. As overseas shipping costs fell during the industrial revolution, it became possible for massive industrial cities to specialise in large-scale production and ship their goods to customers all over the world. The concentration of industry in cities, however, was still largely a product of the fact that it was costly to move goods over land. In a port city, you could bring in inputs, process them into outputs, and ship them back out. If it had been necessary to move intermediate goods well inland for manufacture at any point, costs would have soared, making profitable production impossible.

But transport costs continued falling. Shipping became much cheaper and more efficient. Air freight became an economic possibility. And improvements in trucking and freight rail led to stunning drops in the cost of moving goods over land. And so where once producers had crowded on top of each other in cities to take advantage of specialisation without blowing their budget on transport costs, they now began to spread out: first into the suburbs, then into cheaper regions of the same economy, and then, finally, into vastly cheaper economies abroad. This process facilitated the rapid industrialisation of Asian giants like Japan and Korea, while also speeding along the process of deindustrialisation in expensive markets in America and Europe.

At this point, many seemed to conclude that distance was if not dead then dying, and that continued fracturing of supply chains was likely to continue. What actually seems to have occurred is a bit more interesting. Supply chains have indeed continued fracturing, but distance has reasserted itself in two important ways. First, in the advanced world, agglomerations of the talented individuals who design these products have become increasingly important. And secondly, information technology, which allows for better coordination of production processes, has once again made proximity a relevant concern in manufacturing. It's possible to coordinate a supply chain that's draped across an archpelago of Asian economies. To maximise the return to this chain, however, it's still necessary to keep plants reasonably close together. A plant located in America is too distant from Asia to make much economic sense; transit time to the rest of the supply chain in Asia is sufficiently long, in most cases, as to erode the gains to just-in-time production, or unexpected changes in designs or orders. Changing transportation and communication technologies facilitated a shift in manufacturing to Asia, then reinforced its presence there.

It's not necessary to talk about this as an entirely organic process. Unquestionably, Asian governments aggressively pursued manufacturing and subsidised it heavily, both directly and through advantageous exchange rates. As the story points out, Asia has capitalised on other advantages, as well. Cheap labour is one. More flexible land-use, labour, and environmental rules are another; China can erect a massive operation in no time at all, staffed with compliant labour and with little concern about the impact of the factory on watersheds, air quality, and traffic. Skill supply seems to matter as well. China is churning out engineers with basic technical competence (but less, it appears, than a bachelor's degree) by the hundreds of thousands. It would be incorrect to point to any one of these characteristics as the driving force behind the global shift. Rather, these are self-reinforcing factors within a global economy that has multiple stable equilibria. After some level of Asian development and integration, it became more attractive for manufacturers to locate there as more manufacturers located there.

What does this mean for the American economy? The Times piece quotes Steve Jobs as telling President Obama that those jobs aren't coming back, and they probably aren't. Attracting firms back to America wouldn't simply be a matter of helping reduce production costs in America. You'd have to replicate the convenience of the entire supply chain, which would likely be an enormously costly enterprise. Given the quality of the jobs characteristic of these production chains, one should ask whether it might not be a better idea to invest that money elsewhere.

Apple, it's worth pointing out, continues to capture most of the value added in its products. The most valuable aspects of an iPhone, for instance, are its initial design and engineering, which are done in America. Now, one problem with this dynamic is that as one scales up production of Apple products, there are vastly different employment needs across the supply chain. So, it doesn't take lots more designers and programmers to sell 50m iPhones than it does to sell 10m. You have roughly the same number of brains involved, and much more profit per brain. On the manufacturing side, by contrast, employment soars as scale grows. So as the iPhone becomes more popular, you get huge returns to the ideas produced in Cupertino, and small returns but hundreds of thousands of jobs in China.

This discrepancy manifests itself in America as rising income inequality, which makes Apple's choices somewhat politically fraught. At the same time, it's worth asking how the American government might alter its policies so as to make life better for middle- and low-skill workers in America at reasonable cost. Offering heavy subsidies to Apple to get it to relocate production would reduce inequality in America; you'd increase the tax burden which would mostly affect richer households and you'd create low-wage jobs, which would mostly benefit underemployed, low-skill workers. Now, perhaps after we add up everything Americans will decide that this kind of massive intervention in the economy and associated efficiency cost is worth it, in order to provide the dignity of employment, such as it is, to millions of workers. It's worth asking, however, whether there might not be a different and better way forward.

For one thing, it's far from obvious that the embrace of this system by Asian economies is a good thing for them over the long run. In the short term, as Mr Baldwin points out in his paper, industrialisation via biting off bits of global supply chains is in many ways a more superficial form of development than was achieved in earlier periods by Europe, America, Japan, and Korea, all of which developed an entire production capacity from the ground up. There is far less technology transfer in the newer, supply-chain model, for instance, and so it is less clear that Chinese manufacturing hubs will develop the innovative hubs and headquarters that did emerge in America and Japan. Furthermore, technology can change quite rapidly. It is stunning how quickly the present system emerged it might well go away just as quickly, either through mass automation or changes in input costs or shifts in the cost of moving goods and ideas. Who knows, rapid improvements in 3-d printing could take the world economy back to the days of hyperlocalisation in manufacturing.

America has also done very well in the past by looking forward toward the unseen opportunities ahead rather than the obvious opportunities currently being exploited. Think for a moment about the world Apple has created: one in which a very large share of the population is connected to each other and to all sorts of data resources via remarkable and powerful little hand-held computers. The potential to develop new business models for the production and delivery of goods and services is almost unimaginable, and it's safe to say we've only begun to scratch the surface. One can't say in advance what the impact of such businesses will be on employment and earnings, but it might well change the demand for skills within America in unexpected ways. To the extent that China's government is encouraging resources to flow toward manufacturing and assembly, it is reducing its ability to experiment with and develop these new technologies and businesses.

Jobs matter. They certainly matter to the well-being, material and psychological, of those who struggle to find one. Whether they matter enough—and specifically whether these low-skill manufacturing jobs matter enough—to undertake major and costly government interventions in the economy, in the process potentially harming the effectiveness of America's innovative businesses, is the question with which American workers and leaders are now wrestling. The answer seems clear enough to me, though my perspective is obviously different from those elsewhere in the economy. Until there is a meaningful improvement in American labour markets, however, especially for those without a college degree, recapturing those assembly jobs from China will continue to linger as a policy temptation.