Our labour-intensive factory slump is no small worry

is a Mint columnist and a former Financial Times foreign correspondent.

It has been almost a decade since Harvard economist Dani Rodrik coined the term ‘premature de-industrialization.’ India’s share of manufacturing in GDP may have stagnated for about as long, but with shiny new production-linked incentive schemes making headlines over the past few years, it was possible to believe that India might buck the trend. Rodrik’s thesis was that manufacturing is plateauing or declining as a share of GDP in many developing countries at a relatively early stage in their development.

Recent export data, however, suggests that India is de-industrializing rapidly in labour-intensive sectors. The Indian Express this month quoted the Federation of Indian Export Organisations as saying, “An analysis of sector-wise export performance for the last five years reveals the troubling pattern that India is experiencing a decline in global market share across labour-intensive sectors.” The trade group said that apparel, knitted garments, marine products, plastics, and gems and jewellery had grown at just 1 % to 2%. In fact, during 2023-24, while goods exports contracted by 3%, exports of textiles, leather, gems and jewellery and marine products declined 9%. The reasons are manifold and various, ranging from the difficulties that small firms have had adapting to India’s GST regime and an environment of slowing global trade to a sourcing preference displayed by buyers for Vietnam and Bangladesh because they are part of free trade agreements that we have spent several years negotiating but not joining.

A report earlier this year by Global Trade Research Initiative, a think-tank, contains alarming data-points on India’s decline in the global garments market. It states, “In 2023, China exported $114 billion worth of garments, followed by the EU with $94.4 billion, Vietnam with $81.6 billion, Bangladesh with $43.8 billion, and India with just $14.5 billion… From 2013 to 2023, Bangladesh’s garment exports grew (cumulatively) by 69.6%, Vietnam’s by 81.6%, but India’s grew by only 4.6%.”

This matters because India’s textile and garments industry employs millions. New Delhi, regardless of which party is in power, prefers wooing big-name multinationals such as Tesla while exporters in these sectors tend to be unknown companies. By contrast, China in its early industrialization phase of the 1980s and 1990s leaned heavily on the expertise of Hong Kong and Taiwanese small and medium enterprises that made China’s provinces of Guangdong and Fujian light-manufacturing hubs that then pole-vaulted up the industrial value chain.

There is far better news in our data on exported services, especially over the past decade. These grew from $167 billion in 2013-14 to $340 billion in 2023-24. The growth has slowed in the past year and the worry is that our stubborn stand on data localization, for instance, and more might hobble this golden goose. To some extent, the sheer critical mass of investment and expertise in offshore business process management and analytics protects India from bureaucratic excesses. Just this week, the CEO of Concentrix Corp, a company I had never heard of, told Business Standard that he expects the company’s employees in India to reach 100,000 this year from about 1,000 a little more than a decade ago. I live in Bengaluru and most conversations with anyone in their twenties on where they work are a reminder of this growth we’re witnessing.

The trouble is, as the troubling data on labour-intensive exports shows, India is arguably experiencing the most K-shaped economic recovery anywhere. And the gap with places such as Bangladesh is not limited to their lead in garment exports, but extends to human development indicators, as Swati Narayan observes in Unequal: Why India Lags behind its Neighbours .The book unsettlingly begins with interviewees in Bangladesh laughing at Indians just across the border still defecating in the open. It quotes the National Family Health Survey (covering 2019-2021) which showed that in Bihar, Madhya Pradesh, Uttar Pradesh and Rajasthan, about half of those in villages defecated outdoors. The number of toilets has tripled in Bangladesh in the past three decades. The upshot, as Narayan notes, is that “children in West

Bengal are shorter than Bangladeshi children due to poor sanitation.” Nepal has improved on sanitation too; by 2016, 85% of its homes had toilets and its government was even denying benefits to those who did not have functional toilets. The biggest beneficiaries of widely used toilets and better sanitation, which improves nutritional outcomes, are women and small children.

An Indian-born garment-factory owner, who employs tens of thousands of workers and has invested in Bangladesh for decades, says its wide availability of female garment workers is its chief attribute; a larger share of its girls is in secondary school than in India. Narayan contrasts the life expectancy of a girl born in 2021 in different parts of South Asia: it is 80 in Sri Lanka (whose schooling levels kept pace with Kerala’s for more than a century) and 74 in Bangladesh, versus 69 in India and Pakistan.

Nobody with a cursory interest in economics or a white-collar job would trade our macroeconomic position of ballooning foreign reserves with that of our neighbours. On the ground, however, a woman garment worker in Bangladesh—or even a tourist journeying from the messy sidewalks of Bengaluru to a Colombo of pristine lakes— experiences a different reality.

Stagnation in sectors that employ large numbers can have a direct impact on our basic indicators

Apple has made some of the world’s most iconic advertisements ever. When the brand’s Macintosh computer was launched about four decades ago with its 1984 commercial, whose protagonist was shown throwing a sledgehammer to smash a giant screen depicting ‘Big Brother,’ the world applauded. But when Apple Inc recently released its Crush! commercial for the iPad that showed several things integral to human lives being crushed by a huge industrial compressor, the world shuddered in fear. The predominant view expressed through social media was that ‘Big Brother has returned.

While launching the latest iPad commercial, Apple Inc’s chief executive Tim Cook tweeted, “Just imagine all the things it’ll be used to create.” But instead of putting across the intended message of the device enabling creativity, the commercial ended up conveying a sense of the destruction technology could wreak on all that human hands can create. While watching that commercial, the first thought that went through my head was this: How could Apple have thought it fit to release this advertisement?

Months of strategic discussions happen before an idea for a commercial is finalized. Pre-testing the advertising idea is often done before the final story board goes into production. It is very difficult to believe that at no stage of this long production process did Apple’s marketing department receive any negative feedback on the storyline of this commercial. While the exact reason why Apple released this advertisement is hard to understand, one thing is clear. The public response to it should serve as a signal for the artificial intelligence (AI) industry.

For all the good things AI machines bring into this world, the fear that these modern contraptions will take over the world and all that is human could one day be subservient to them is an anxiety that has expanded its presence in the world’s non-conscious. The reaction to the iPad commercial was immediate and widespread, just as we would expect if the world were to suddenly realize that all that’s human will soon be replaced by machines.

In these times of social media, such intense fears spread like a wildfire across the world. If such non-conscious fears of what AI will do to human roles persist, it could lead to even stronger pushback against the global AI industry.

This resistance is unlikely to take the form of a few protests here and there, displays of anger where a few technology products are destroyed in public places, as Luddites did to spinning mills at the start of the Industrial Revolution. The pushback could be significant enough for the AI industry to stay on guard.

Protests will probably aim for the Achilles heel of the industry. Questions have arisen on the privacy of the vast data used by the AI industry to train its models, as also over the ownership of this intellectual property. The huge carbon footprint of AI training is another target of critics, just as job losses caused by AI has been flagged as a major problem. These are issues on which the AI industry is yet not in a strong position to defend itself. Rather, it is on the back-foot on all these. A pushback on any of these issues could slow down an industry that has shown tremendous growth momentum in recent years. This is why AI players across the world should be careful about the language they speak and positions they take.

Let’s take the example of an AI product developed to read X-ray images and detect signs of tuberculosis. How should such a innovative AI product be positioned? It can be positioned as a tool that can read X-ray photos with far greater accuracy than a human radiologist. One could easily work out the productivity gains a hospital would make by automating this task. But this positioning stance has a weakness: the AI product would be portrayed as heroic, while the human professional will get cast as inadequate.

The same AI product can also be positioned as a product that helps detect tuberculosis among people who have no access to a radiologist. This positioning is all about meeting an unmet human need, an endgame where humans emerge as winners. There is no scope for doubt which of these two approaches taken to the market will win public approval and which is likely to run into resistance.

AI products that try to complement human roles instead of replacing them will be welcomed. This calls for a deeper understanding of need-gaps in existing products and developing innovative products to fulfil those unmet needs.

It is easy to develop a machine that fits into an air-conditioned room of a city hospital, but a machine that can help read X-rays of people in remote villages located at high altitudes in the Himalayan mountains, or in rural settlements of the Thar desert, speaks of innovation driven by human needs.

Recent turns in history remind us that humans have not let inventions flourish that are seen as detrimental to human existence. Take the case of nuclear energy. It offered humanity the benefit of power generation free of fossil-fuel emissions. But when people realized its larger harmful effects and risks, public rejection and legislation followed that have kept the nuclear-power industry under stiff restrictions.

The iPad advertisement’s backlash should remind the AI industry to get its language right. The pushback is real, and if AI is seen as a threat to humanity, it will only grow.