The world is in the grip of a manufacturing delusion

In the West the aim is to reverse industrial decline, which is keenly felt by voters. As a share of global economic output, manufacturing has dropped from 19% in 1997 to 16% today, with the fall steepest in rich countries. In China and India industry’s share of economic output appears to be roughly where it was three decades ago, but even in these countries it has slipped in recent years (see chart 1).

Manufacturing boosters make four arguments in favour of attempting to reverse this trend. First, politicians in the West say that factories are a source of solid jobs that produce a bigger and more satisfied middle class. Second, boosters view manufacturing as a driver of innovation and growth. This is urgently needed to fuel the green transition—the third reason—which will be more palatable to electorates if it delivers local jobs. Finally, tensions between America and China have pushed world leaders to reconsider which goods are strategically important, and therefore should be produced closer to home.

Start by considering the type of employment on offer. The notion of a “good manufacturing job” is an old one. During the 20th century, those without a university education could find decent wages, job security, a bit of personal autonomy and career progression in factories. Indeed, just over a decade ago production jobs in America paid a premium of 5% compared with similar service-sector ones, and offered steady hours and generous benefits.

More recently, though, the picture has changed: many good manufacturing jobs no longer exist (see chart 2 ). Across the rich world, employment that requires mid-level technical skills (think machine operators) has given way to a mix of high- and low-level jobs, mostly in service sectors (think coders and baristas). Wealthy populations spend more of their income on services; industrial demand is increasingly met by emerging markets. Whether in Detroit or Dortmund the consequences are obvious and, by now, familiar. At first glance, those who decry the loss of manufacturing jobs have a point. Surely it is worth paying to get them back?

There is a snag. It is far from clear such jobs can be brought back—no matter how much governments spend. For a start, the manufacturing wage premium has fallen sharply. Production workers’ wages in America now lag behind those of similar service-sector workers by 5%. Moreover, the sort of high-tech factories that America and Europe are attempting to attract are highly automated, meaning they are no longer a significant source of employment for people with few qualifications.

Ford’s revamped electric-vehicle (ev) plant in Cologne, located on the banks of the Rhine in Germany’s industrial heartland, is one such example. The chassis and bodies of vehicles are coated in chemicals to prepare for painting and to prevent corrosion. This happens across multiple storeys; the number of workers involved in the work on site is precisely zero (two keep tabs remotely). Shiny yellow assembly robots further down the production line are sufficiently advanced as to be able to mostly monitor themselves. Although workers are required for assembly—about as many as for traditional petrol-powered vehicles—the activity requires a lot more training. This matches the national picture: according to a study by Wolfgang Dauth of the Institute for Employment Research and co-authors, industrial robots have made available work more complex.

Economic historians are also increasingly questioning whether state support for manufacturing did, in fact, play as decisive role in the economic development of East Asia and the West as is commonly assumed. At the very least, productivity growth in services and the removal of protectionism was also crucial. Even if industrial policy was perfectly designed by a clear-sighted government, which used tax and subsidies to shift labour to industries with the greatest economies of scale, it would only bring a one-off 1-3% boost to gdp, according to a paper by Dominick Bartelme of the University of Michigan and co-authors. Indeed, in recent decades there has been next to no relationship between economic growth and manufacturing’s share of the economy among countries in the oecd (see chart 3).

The oft-lauded superior productivity growth of manufacturing—versus services as well as agriculture—comes with caveats. Economists have found that financial, it and legal services can boost productivity elsewhere, including in industry. According to the imf, the gap between manufacturing and services productivity growth has shrunk in many countries since the turn of the millennium. In China and India its direction has flipped, with services productivity rising faster. Moreover, services are a broad church, ranging from teaching to tech. The latter boasts extremely fast productivity growth, which may soon be propelled further by artificial intelligence.

Industry’s greater capacity for innovation also comes with something of a catch. “Measuring innovation spending is easier in manufacturing firms, [which tend to] have dedicated research-and-development (r&d) departments,” notes Paul Hünermund of the Copenhagen Business School. State spending on r&d is therefore often directed at industry, he explains, even though services may be more innovative than typical measures suggest.

The strongest argument for manufacturing’s importance to innovation comes from economists who point out that lots of tech progress happens in places where industry and services co-exist. Gary Pisano and Willy Shih, both of Harvard Business School, stress the importance of learning-by-producing in industries such as high-tech chipmaking. The idea has been embraced by politicians, including in America, China and Germany. Yet even here evidence is mixed. In 2001 Taiwan eased an offshoring ban, resulting in some production moving to China. Academics who studied the consequences found that innovation in the affected goods did decrease, as the co-existence thesis suggests. It was, though, balanced by a rise in innovation in other goods and tech, since r&d resources were freed up. The result was specialisation, not lower overall innovation.

The same is not true for other green equipment, where demand will create supply. Solar panels show how the process will probably play out. The current boom in demand—America installed 47% more in the first quarter of 2023 than in the same quarter last year—has prompted companies in China and elsewhere to boost capacity. The Energy Transitions Commission reckons that existing production of solar panels already exceeds probable demand until 2030, which is also the case for batteries when planned production is included (see chart 4). In other areas, like heat pumps, capacity can be added quickly if desired.

Politicians also hope that spending will create new firms and even industries. Tesla and byd, the American and Chinese upstarts of the ev age, are the sort of thing they have in mind. Energy-intensive factories in places with abundant, and cheap, renewable energy would also be nice. But from a global perspective, a subsidies race is ultimately zero-sum. The gigafactory attracted to France after the government offered ProLogium, a Taiwanese firm, sweeteners, cannot also be built in the Netherlands. Nor are countries likely to reap benefits that go much beyond the amounts they splash out. Reports suggest that Germany has increased its offer to Intel from roughly €7bn to €10bn in an attempt to entice the chipmaker to build a new fab in Magdeburg, two hours from Berlin. Although the region will benefit through jobs and visitors, Germany is unlikely to be better off, since taxpayers will have funded most of the wider benefits.

Governments have long protected military manufacturing in order to ensure they have plentiful supplies of weapons when required. But the list of goods for which countries do not want to depend on others has grown in recent years—the final argument for spending on industry. Supply disruptions during and after the covid-19 pandemic demonstrated widespread vulnerabilities: a scramble for protective medical equipment bolstered the case for stockpiling; a shortage of usually common chips slowed car production; a lack of gas in Europe after Russia curtailed supply provided an alarming example of how economic dependence could be exploited.