China as number three?

OECD long-term scenarios, released without much fanfare in September last year, suggest that India will become by 2060 the largest economy in 2021 USD purchasing power parity terms (PPP). By the end of this century, India will be almost twice the size of China (see Figure 1). The US, now in second place, will slip to third in the 2040s, but then catch up with China again in the 2070s, two centuries after it first overtook China to become the world’s largest economy in nominal terms in the 1870s. In previous OECD scenarios released in 2023, the projection horizon was 2060, and China was projected to still be the largest economy by then by a considerable margin. This is still just the case in the current projections, but the significant changes occur in the second part of this century.

This is largely because of demographics. The United Nations’ Population Division projections released in 2024 suggest that China’s population will decline much faster than had been expected, with most of the decline in the second half of the century.  The median projection puts China’s population at some 650 million people by 2100 from about 1.4 billion today. India will have the largest population in the world by far in 2100, at 1.46 billion, little changed from about 1.43 billion now, after having peaked at 1.65 billion in 2060.    

Even with a projected rise in labour force participation in terms of percentage of the population, China’s labour force will be less than 300 million by the year 2100, from 734 million now, which is barely 40% of that of India’s and not that much larger than the US’s (see Figure 2). India’s labour force participation is expected to increase rapidly from some 40% of the working age population (15–74 in the OECD projection) to some 70% by 2100. The US labour force, due to liberal immigration policies, also continues to expand, and is a key driver behind the US overtaking China again. However, it is not only population dynamics that explain India’s surge and China’s lagging behind.

The OECD uses a fairly standard model for generating its scenarios. It uses a Cobb-Douglas function with constant returns to scale, and two production factors: labour and capital, and global technical progress. Basically, the model predicts countries’ gross domestic product (GDP) to converge, but the rate of convergence depends on the change in “labour efficiency” and the initial capital/labour ratio. In a steady state, all countries would grow with the global rate of technical progress, but most of them are catching up with more advanced ones. The rate of change in labour efficiency depends on an individual country’s institutional strength and its distance to the efficiency frontier. This is highly stylised in the model and measured by three factors: (historic) macroeconomic stability; rule of law; and economic openness, which no doubt is somewhat arbitrary.  Regardless, the values for changes in labour efficiency are about the same for China and India, but India’s efficiency of labour grows faster simply because it is further away from the frontier.

Another factor in convergence is capital stock dynamics.  Due to the assumptions of the model (Cobb-Douglass, labour share of 67%), the ultimate capital output ratio is a given at 3.4, according to the OECD. But since China has more capital per worker to start off with (and sees the number of workers decline over time, which increases the capital labour ratio), China gets less GDP growth than India from the same level of investment.

Together, these factors explain why China’s productivity growth is lower than India’s (see Figure 3). Productivity here is simply calculated (by the author) as GDP divided by the workforce. But this is not a given. For instance, China could grow faster than the OECD scenario by “beating” the catch-up rule in labour efficiency, say by applying artificial (AI) intelligence and robotics faster than India and faster than the “frontier” country, the US. China may also grow faster if the income share of labour stays below the level imposed by the model, which seems to be the case currently. China’s current policy orientation to boost consumption, however, would suggest that the labour share will rise over time.  

The OECD scenarios are not cast in stone, and are not destiny as “long-run scenarios should not be treated as predictions of the future but rather as illustrations of some of the long-run challenges facing the global economy, how they might evolve, and how they might affect different countries”. Further, technological innovation in AI or other technologies could bring about massive changes in productivity, possibly dramatically changing the underlying assumptions for the scenarios. Nevertheless, demographics play a big part, and India’s projected margin over China in size of GDP by the end of the century is very large, so it seems safe to say India will be the largest economy by then.

This article was first published as East Asian Institute Commentary No. 98.

Related: As US tariffs bite, China and India need each other more than ever | Can India move beyond faltering US ties?