A report from the Boston Consulting Group last week suggested the U.S. had become the second-most-competitive manufacturing location among the 25 largest manufacturing exporters worldwide. While that news is welcome, most of the lost U.S. manufacturing jobs in recent decades aren’t coming back. In 1970, more than a quarter of U.S. employees worked in manufacturing. By 2010, only one in 10 did.
The growth in imports from China had a role in that decline–contributing, perhaps, to as much as one-quarter of the employment drop-off from 1991 to 2007, according to an analysis by David Autor and colleagues at the Massachusetts Institute of Technology. But the U.S. jobs slide began well before China’s rise as a manufacturing power. And manufacturing employment is falling almost everywhere, including in China. The phenomenon is driven by technology, and there’s reason to think developing countries are going to follow a different path to wealth than the U.S. did—one that involves a lot more jobs in the services sector.
Pretty much every economy around the world has a low or declining share of manufacturing jobs. According to OECD data, the U.K. and Australia have seen their share of manufacturing drop by around two-thirds since 1971. Germany’s share halved, and manufacturing’s contribution to gross domestic product there fell from 30 percent in 1980 to 22 percent today. In South Korea, a late industrializer and exemplar of miracle growth, the manufacturing share of employment rose from 13 percent in 1970 to 28 percent in 1991; it’s fallen to 17 percent today.
The decline in manufacturing jobs isn’t confined to the (now) rich world. According to the Groningen Growth and Development Center, manufacturing jobs in Brazil climbed as a proportion of total employment from 12 percent in 1950 to 16 percent in 1986. Since then it’s slid to around 13 percent. In India, manufacturing accounted for 10 percent of employment in 1960, rising to 13 percent in 2002 before the level began to fall. China’s manufacturing employment share peaked at around 15 percent in the mid-1990s and has generally remained below that level since, estimates Harvard economist Dani Rodrik. As a proportion of output, manufacturing accounted for 40 percent of Chinese GDP in 1980 compared with 32 percent now.
As Rodrik has pointed out, most of today’s rich countries “became what they are by traveling the well-worn path of industrialization.” Agricultural workers moved to factories, followed by a period when manufacturing ceded its dominance to services. Because of the declining demand for labor in manufacturing, however, the traditional path from peasant through factory worker to cubicle farmer is missing its middle step. That’s particularly bad news, suggests Rodrik, because, unlike labor productivity in general, productivity in manufacturing is converging across countries—it’s rising faster in countries that start with the lowest productivity. And the global decline of manufacturing could have political effects: Manufacturing unions were an important part of the organization of labor parties worldwide.
The picture isn’t all grim. First, manufacturing alone was never sufficient to drive a country from poverty to wealth. A common feature of advanced economies is that the share of output and employment in each sector is broadly the same: In high-income OECD countries, industry accounts for 24 percent of value added and 22 percent of employment. Agriculture accounts for 1 percent of value added and 3 percent of employment (while that’s a large relative gap, the absolute difference is small). Every sector has “industrialized,” and labor productivity is broadly similar across them. In developing countries, by contrast, agriculture accounts for 11 percent of output but fully 38 percent of employment, while the reverse is true in industry and services, where there’s a larger share of output than employment. The rebalancing of jobs across major sectors doesn’t only involve manufacturing growth—the whole economy must become much more efficient and capital-intensive.
The importance of nonmanufacturing sectors to the growth story may help explain recent global economic performance. It’s true that the last 10 years have been bleak in developed countries. But, for all the decline in manufacturing job share, the developing world has had a fantastic decade of economic growth driven in large part by a stronger services sector. Services accounted for about one-half of global output in 1980, now they make up 70 percent. In Brazil, the share is 67 percent; in India, 55 percent; and even in China, where manufacturing remains outsize, services, at 43 percent, still account for a larger proportion of GDP.
Worldwide, services account for 70 percent of value added and 45 percent of employment—the sector “outperforms” in terms of labor productivity. That’s true in some developing countries, too. The most recent Groningen data for India suggests manufacturing accounted for 11 percent of employment and 16 percent of value added. In the same year, services accounted for 22 percent of employment and 49 percent of value added. In other words, services employ more people than manufacturing and average value added per person is a lot higher.
In China, value added per worker is higher in manufacturing than in services, and there’s huge variation across different parts of the services sector. In some areas, including the small kiosks run by micro-entrepreneurs who sell a few items a day for lack of better opportunities, productivity is incredibly low.
Across the developing world, there are small manufacturing operators in similar dead-end occupations. Jobs in Bangladesh’s garment industry may be better than the alternatives faced by low-skilled workers in that country, but one year after the Rana Plaza garment factory collapse, which killed more than 1,000 people paid about $40 a month each, we shouldn’t romanticize factory jobs. In India, nearly two-fifths of employees in manufacturing work in textiles, leather, food, beverage, and tobacco industries where compensation is less than $1 an hour. That compares with the one-fifth working in the machinery, vehicles, and computer-and-electrical-equipment sectors, where compensation reaches above the princely sum of $2 an hour. U.S. Bureau of Labor Statistics data suggest the average manufacturing employee in India or China in 2009 was compensated at about 5 percent of the amount earned by her U.S. counterpart. Adjust that for the different purchasing power of a dollar in each country, and Indian manufacturing employees are earning about an eighth of what their U.S. counterparts make.
Many basic service jobs do better than that. Orley Ashenfelter, writing for the U.S. National Bureau of Economic Research, calculated how much workers in McDonald’s franchises around the world are paid in terms of Big Macs an hour (a way to adjust for differences in purchasing power). Chinese workers (at 0.6 Big Macs an hour) are paid about one-quarter of the U.S. rate, while Indian workers are compensated at about one-seventh the rate. That’s better than the average manufacturing job, and the working conditions are probably better, too.
Subsectors such as food retail are undergoing dramatic change worldwide. In Brazil, supermarkets had a 75 percent share of retail food sales in 2003. That number is only 5 percent in India, but the sector has been growing 49 percent a year recently, according to researchers at the International Food Policy Research Institute. Mobile communications is one part of the services sector where growth in productivity has exploded because of the spread of technology. The number of mobile subscriptions worldwide reached 6.8 billion in 2013. In India alone, the mobile industry was estimated to employ about 2.8 million people in 2012.
Leapfrogging the factory to the checkout counter and cubicle farm appears to be far more straightforward than it might once have been. Compared with Western countries when they were undergoing the transition from agricultural employment, developing countries have workforces that are healthier and better educated, with access to technologies including mobile phones and computers that didn’t exist when manufacturing seemed the only way to climb the productivity ladder.
And while factories have been the traditional base of political organization for those at the lower end of the income distribution, the 2.1 million members of the Service Employees International Union in the U.S. and the 1.4 million members of the Teamsters union suggest manufacturing isn’t the only sector that can organize. Bureau of Labor Statistics data on private-sector employment suggest only 11 percent of U.S. manufacturing employees are unionized, compared with 22 percent in transport and 15 percent in education, telecommunications, and construction. (Total unionization in the U.S. was almost 35 percent in the mid-1950s.) Regardless, many developing countries haven’t needed large manufacturing unions to push labor regulation. Indeed, in India, businesses complain that regulations are far too generous to labor. In 2011, if a firm had more than 100 employees, it could sack a worker only for criminal misconduct.
Those bemoaning the decline of manufacturing are right that the trend means the traditional path to economic development no longer works, and they’re right that we’re not sure yet what the new path looks like. So the conventional wisdom on growth strategies probably isn’t much use. Frankly, that’s pretty much always been the case: The conventional wisdom in the 1960s was that Ghana would grow more rapidly than South Korea, and the conventional wisdom in 1980 was that China had terrible prospects for economic development while Japan would go on growing forever. What we do know is that more people worldwide are finding good jobs outside the manufacturing sector these days, and growth in the developing world has sped up over the past 20 years despite the decline in global manufacturing. In the U.S., India, and China alike, there’s life beyond the factory gate.