Bernstein, a major Wall Street research firm, has just published an open letter to Indian Prime Minister Narendra Modi laying out eight reforms it argues India must complete before the country's growth story stalls. The letter is unusually blunt for a financial document. It warns that despite India's recent achievements — including overtaking Japan as the world's fourth-largest economy — the country is at real risk of under-delivering on its potential. The next decade, Bernstein concludes, will decide whether India's rise is real or merely an artefact of favourable comparisons.
The headline worry concerns the labour market. Roughly ten to fifteen million Indians work in information-technology services, call centres, and back-office roles for global companies. Generative AI can now perform a substantial share of that work. Meanwhile, the high-value parts of the AI revolution — the foundation models, the chip designs, the patents — sit overwhelmingly in the United States and China. The risk Bernstein flags is that India becomes a customer of the AI revolution rather than an owner of it.
The structural numbers reveal a deeper imbalance. About forty-two to forty-five percent of India's workforce still works in agriculture, but farming generates only fifteen to sixteen percent of national output. India imports roughly eighty-eight percent of its crude oil, which makes energy prices a national-security issue rather than a purely economic one. State-run electricity distributors are sitting on accumulated losses of more than five trillion rupees, somewhere between sixty and seventy billion dollars, which industrial consumers effectively subsidise through higher tariffs. Each of these is a brake on the kind of factory-led growth that lifted East Asian economies out of poverty.
The 'middle-income trap' is the term economists use for what Bernstein is describing. The easy gains from cheap labour begin to run out, but the country has not yet built the innovation infrastructure to compete with rich economies. The relevant comparison is not India today versus India twenty years ago, but South Korea in 1990 versus the Philippines in 1990. Both countries had similar incomes at that point. South Korea bet on building Samsung and Hyundai. The Philippines stayed in services. Three decades later, South Korea is roughly four times richer per person.
The 'China plus one' strategy, in which global firms diversify supply chains away from China, has been slower to benefit India than headlines suggested, because building factories requires reliable power, ports, and supplier networks. Three signals will tell the story of the next decade. Whether private companies actually build factories rather than merely announcing them. Whether power-sector reform finally happens. And whether any AI model designed and trained in India breaks into the global top tier. If two of those three fail, the question Bernstein poses — does India build innovators or task-rabbits — will answer itself.
India just overtook Japan to become the world's fourth-largest economy — but a leaked Bernstein letter to its Prime Minister warns the next decade could decide whether that rise is real or a mirage.
Bernstein, a major Wall Street research firm, just published an open letter to Indian PM Narendra Modi laying out eight things India must fix before its current growth story stalls. The letter is unusually blunt for a financial document — it argues India is at risk of "under-delivering" on its potential despite recent wins.
The headline worry: roughly 10–15 million Indians work in IT services, call centres, and back-office roles for global companies. Generative AI can now do a lot of that work. Meanwhile, the high-value parts of AI — the models, the chips, the patents — sit overwhelmingly in the US and China. India risks becoming a customer of the AI revolution rather than an owner of it.
Every developing country eventually hits what economists call the "middle-income trap": the easy gains from cheap labour run out, but the country hasn't yet built the innovation muscle to compete with rich economies. India is approaching that fork in the road right now.
India will add more people to the global workforce this decade than any other country. If you're choosing a college major, watching the AI job-market debate, or wondering why every tech company suddenly has a Bengaluru office, this is the backdrop. The same question Bernstein asks India — do we build innovators or task-rabbits? — is the question every economy facing AI is quietly asking, including the US.
History rhymes here. Japan in the 1960s, South Korea in the 1980s and China in the 2000s all faced a "now or never" decade where policy choices about education, energy and manufacturing locked in their fates. India's window is open but narrowing. Watch three signals over the next few years: whether private companies actually build factories (not just announce them), whether power-sector reform finally happens, and whether any India-built AI model breaks into the global top tier. If two of those three fail, the engineers-vs-delivery-drivers question answers itself.