Industrial policy is back, but the implementation challenge remains

The return of industrial policy to the top of the global development agenda has been formalised with a new World Bank report on the topic. The summary is here. I blogged earlier on this here. 

Notwithstanding the standout success of the East Asian economies in using industrial policy to drive economic growth, for long the WB and others had sought to minimise its role and explain it away by pointing to their unique circumstances. Specifically, their high educational attainment, high savings rate, low inequality, and the capability of the governments to implement such policies. At best, they were targeted to address specific problems in the functioning of markets. These conditions, it was said, did not exist in other countries. 

The report admits that industrial policy is now more replicable. It points to increased capabilities due to higher education levels, political support for delivering development objectives, and greater global trade integration. On their own, none of these is compelling enough for a re-examination. It is most likely that the widespread adoption of industrial policy by developed countries in response to Chinese competition is behind this change of mind within the Bank. 

Two comments in this context. It would have been more appropriate for the WB to acknowledge that its assessment was incorrect and to revise its views. Two, given their own market dominance in manufacturing, it was in the interests of developed countries to then oppose industrial policy in developing countries. Now that China has assumed dominance, it has reverted to the same set of policies that allowed it to become dominant in the first place. 

The WB report defines industrial policy as “the range of policy tools that governments use to shape what an economy produces rather than leave it to the discretion of markets alone”. It says that upper-middle-income countries now spend a record 4.2% of GDP on business subsidies. Low-income countries target growth of 13 industries on average, more than twice that in high-income countries. 

It identifies 15 industrial policy tools and categorises them, as in the table below. 

It analyses the minimum requirements for their adoption. The cells marked large indicate the tool needs a large fiscal space, market size, or government bandwidth, and vice versa for those marked small. 

In general, when government capacity is small, local market size is small, and fiscal space is also small (as in the case of most low-income economies still), industrial parks constitute the best industrial policy tool. When all three of those advantages are large (as in the case of the world’s biggest economies), then import tariffs, innovation subsidies, local content requirements, and many other policy tools can be effective... These countries enjoy the widest range of opportunities to experiment with industrial policy for development… this report finds that low-income economies—usually characterized by small market size—tend to be the heaviest users of import tariffs, which require a large market size to be effective. Upper-middle-income countries, for their part, go big on business subsidies.

The report has the latest glowing impact assessment on one of South Korea’s industrial policy decisions, 33 years later.

They found that the impact of the government’s big push for heavy and chemical industry (in the 1970s) caused the economy’s GDP to be 3 percent larger each year in the long run. This benefit far exceeds the economic cost of the government’s “large subsidies,” estimated by the World Bank’s 1993 report at 2.4 percent of GDP in only one year.

While all 15 policy tools are well known, the report provides a very broad but widely-known framework for understanding and applying industrial policy, and has a good compilation of case studies. Beyond this, I’m not sure how relevant or useful the report is for policymakers and implementers struggling to make decisions on their industrial policy. 

The framework presented in the report can perhaps be reduced to three different instruments of industrial policy - regulatory restrictions (restrict access to foreign firms/investors/imports), regulatory enablers (preferential treatment to domestic firms/investors), and fiscal support (tax concessions, input subsidies, capex incentives). 

In theory, mandates (regulatory tools) appear simpler than subsidies since they seek to directly target the required change. For example, domestic content or technology transfer mandates or commodity export bans (Indonesia’s ban on nickel and now bauxite exports) are easier to adopt compared to subsidies for production, innovation, consumption, or exports. But mandates create the risk of suboptimal outcomes - poor quality, capture by vested interests, constraints on production (e.g., due to limited supply), poor uptake, etc. Compliance with the mandates becomes even more of a challenge when the manufacturing supply chain is so utterly dominated by China.

An underlying assumption behind advisorial policy prescriptions by entities like the World Bank, consulting firms, or research experts is that they are actionable and can be readily applied by practitioners without much effort and existing capabilities. This assumption is most often incorrect.

Assuming the objective (say, building local manufacturing capabilities) and the target industry (say, electronics devices), a practitioner who is preparing a proposal for financial appraisal and subsequent approval by the competent authority generally struggles with the following five questions.

1. What combination of instruments is adequate to achieve the objective?

2. What should be the degree or extent of incentivisation or regulation? Specifically, how much is just about enough to trigger the market response? More likely, is the proposed measure sufficient to trigger the supply response?

3. What should be the implementation design of the chosen policy tool (especially relevant for tools like skill development, market access assistance, innovation subsidies, R&D tax credits, etc.)? 

4. How to know that the industrial policy is having its expected impact, and what revisions and course corrections are required?

5. What are the likely distortions from these interventions, including abuses during implementation, and how can they be mitigated?

The policymaker or implementer must make choices in constrained conditions - information, time, resources (financial and physical), acceptance, and capabilities. 

For one, there is the absence of good data to assess, prioritise, and make choices. Second, the theories of change are complicated, especially given the multiple confounding factors. Third, the problem-solving required to figure out the combination of choices is often a challenging exercise that does not get done well. Fourth, given fiscal constraints, the sponsoring departments often end up being forced to compromise substantively on the unit rates for any fiscal support (like the percentage of concession or extent of subsidy). The fiscal support ends up being below the threshold and fails to generate the market response. Fifth, there are the deficient state capabilities that weaken effective implementation. Sixth, there’s no way to know whether the policy combination is meeting the intended objectives, beyond some headline numbers which are often deceptive and misleading. And so on. 

To illustrate the points made above, the table below captures the uncertain elements. The tools and frameworks outlined in the WB report cannot be of much help in resolving the uncertainties and making decisions. The details of these elements matter, and are generally the difference between success and failure. Getting these details right is hard. 

In this context, there may be some useful principles for resolving the aforementioned uncertainties and formulating an industrial policy that is likely to succeed. 

In responding to a policy problem, it is advisable to start with a combination of instruments that are identified and prioritised using internal institutional judgment (and this is the way it happens in practice). But there should be rigorous monitoring systems that use administrative data (and targeted rigorous surveys, as required) to generate feedback on performance and impact, mainly to assess the directionality and order of magnitude of impact. Almost always, it is the absence or inadequacy of the latter that is a far bigger problem than the rigour of the choice and design of the former (combination of instruments). Surprisingly, the report overlooks this aspect completely. 

It is typically the case that the extent of subsidy depends on the fiscal envelope available and/or the priority accorded to the objective or target industry by the government. If the primary objective is a national priority, then compromising on the unit rates (or the extent of subsidy) should be strictly avoided. The same risk is there with mandates - how much restriction is sufficient? The feedback loop mentioned above is critical for getting the details of both subsidies and mandates right.

Once the higher-level choices are made, it is essential to avoid constraining implementation with onerous eligibility, access, documentation, compliance, and general procedural requirements that are typically put in place to eliminate abuse and fraud. There should be less invasive and onerous mechanisms to achieve the objective of abuse elimination. Further, the system should condone some level of slack or abuse, like false positives (Type I errors).

Most importantly, competition is critical to ensure that industrial policy is not captured by vested interests. Just as the performers should be rewarded, the non-performers should be progressively excluded. As Joe Studwell has written, export competition was critical to the success of the Northeast Asian economies. Other arrangements/metrics of competition are therefore critical for the success of industrial policy. It shapes the overall expectations and aligns incentives for the policy to play out for the long-term.

Weekend reading links

1. China may be turning to its traditional industrial policy playbook to reduce its current import dependence and achieve self-sufficiency in food, especially animal protein, thereby upending global supply chains. 

Adam Tooze writes, pointing to analysis by the research group Systemiq.

This involves a combination of investment and innovation. Beijing is coordinating central and provincial government, state-owned enterprises and financial institutions around smart agriculture. It has licensed the commercialisation of genetically modified maize and soya. Research clusters are forming around neoproteins, fermentation-derived ingredients, feed additives and agricultural biotechnology. State banks are on hand to provide cheap finance. To channel demand, Beijing is tightening food and feed standards and tweaking procurement requirements. This is the kind of whole-system policy that has given China a commanding lead in the new energy sectors. With all the levers in play, we may, by as early as 2030, see a significant fall in soyabean demand, slashing imports from the US dramatically. By 2040, innovation and efficiency gains could plausibly turn China into a net exporter of poultry, dairy, eggs, fish and seafood. If agriculture follows the industrial policy timeline, by 2050 we should expect to see China emerging as a major source of “cultivated meat”.

2. Indonesia's President Prabowo Subianto has signed a presidential regulation cutting the percentage that ride-hailing platforms like Grab and GoTo can take from each order to a maximum on 8%, down from the current 20%. 

3. The latest on the dominance of the Magnificent Seven (US stocks make up 65% of global stock market capitalisation, up from 40% in 2010).

This is an important data point.

Bloomberg data suggests that almost 80 per cent of the S&P 500 companies that reported first-quarter results this month beat analyst earnings estimates.

4. Chinese ports facts of the week.

Chinese firms now operate or have a financial stake in at least 129 ports outside China (see map), and have spent at least $80bn on port construction from Antigua to Tanzania, with many of the investments tied to bilateral trade and regional shipping agreements. More than a third of China’s overseas ports are near maritime chokepoints, including the Strait of Malacca, the Strait of Hormuz and the Suez Canal, making them indispensable operators in strategic areas. China’s firm grip on global ports has rattled Western governments. MERICS, a think-tank in Berlin, found that after a terminal operating contract is signed, total trade with China rises by more than a fifth, while countries that allow Chinese firms to run all their terminals at one of their ports see a 19% drop in exports to the rest of the world. Operating ports allowed Chinese firms to prioritise their cargo and vessels and speed up customs and logistics...

China’s reach extends beyond physical infrastructure. LOGINK, a Chinese government-run logistics-management software, is used in at least 24 countries and 86 ports (America banned its use in 2023). LOGINK shares data with CargoSmart, another shipping-management software firm owned by COSCO, another Chinese state-owned firm, and in turn gives it access to the whereabouts of 90% of the world’s container ships. It also has a tie-up with CaiNiao, a logistics provider with hundreds of warehouses around the world... Chinese firms are also building industrial parks and manufacturing facilities close to their existing ports in Africa and Europe.

5. Good FT article on how data centres construction is triggering local backlash as residents fear for their water and electricity.

Around two-fifths of all US data centres are located in areas of high water stress, according to S&P Global... In DeKalb, a city of around 40,000 people, water demand averages just over 3mn gallons a day, rising to in excess of 4.5mn gallons at peak. The latter figure is broadly comparable to the needs of a single large AI data centre.

Data centres have already emerged as a significant driver of economic expansion in the US, accounting for 80 per cent of private sector growth in the first half of 2025, according to S&P Global... Researchers at the Lawrence Berkeley Laboratory forecast that so-called hyperscaler data centres will consume anywhere between 60bn and 124bn litres (16bn and 33bn gallons) of water on-site each year in 2028. This figure excludes the indirect water use tied to electricity generation, which the lab previously forecast could be as much as 12 times higher than direct consumption...

Of the roughly 100GW of additional electricity capacity that the US is projected to need at peak times by 2030, roughly half will be used by data centres, according to the Department of Energy... On average, American bill payers — including residential, commercial and industrial customers — paid over 6 per cent more for electricity year on year at the end of 2025. This increase was starkly higher in the mid-Atlantic states which house a large number of data centres such as Pennsylvania and Virginia, where bills rose by 19 and 10 per cent respectively.

6. Indian equities suffer from a high valuation gap with EM peers.

In recent weeks, a string of global brokerages – including Goldman Sachs, Nomura, HSBC, UBS, and JPMorgan – have downgraded Indian equities in their emerging market portfolios. The concerns converge around a common theme: Deteriorating macro conditions amid rising energy prices, weakening earnings visibility and, crucially, more attractive opportunities in other EMs... The relative valuation gap still stands at around 65 per cent—well above South Korea, Brazil, China, Mexico and South Africa. “South Korea is expected to deliver about 180 per cent earnings growth at around 7x P/E. Taiwan offers about 35 per cent growth at 18x. China delivers about 14 per cent at 11x. India, by contrast, offers 8–14 per cent growth but trades at ~19x,” said Amish Shah, head of India Research, BofA Securities. From a global allocator’s perspective, India remains expensive.

The weight of Indian equities in the MSCI EM index has continued to fall since March 2025, from 18.5% to 12.58% in March 2026.

7. India's trade deficit with China continues to grow.

On April 8, Xu Feihong, Chinese ambassador to India, posted on X: “Glad to know that China has become India’s largest trading partner in FY2026 — for the 11th straight month.” This was great news for China but not for India. The $151.1 billion trade between the two comprised $131.63 billion of exports from China to India and a meagre $19.47 billion of imports from India. It was a one-way street. China was flooding Indian markets even as it became India’s “largest trading partner”...

India must import electronics and electrical equipment ($40 billion-50 billion), machinery ($27 billion), organic chemicals ($12 billion-13 billion), plastics, steel, medical equipment, and so on from China every year. These are critical products without which the Indian economy would not be able to function. Alternative sources exist for some, but at a much steeper price. Moreover, China has diversified its supply sources, so even if not directly from it, Chinese goods would still reach India through Southeast Asian and other manufacturing bases.

8. One explanation with oil markets are still not rising as much as expected. 

In 1973, the year of the first oil price shock, I calculate from industry sources that about 80 per cent of one barrel of oil was consumed per $1,000 of global GDP in 2025 prices — 131 litres, to be precise. In 1980, the year after the Iranian revolution, this was down to 116 litres. Last year, it was 52 litres. The current level is still a lot, but the average oil burden is 60 per cent less than 50 years ago. If so much less oil is needed, real prices should have much more room to escalate before they cause the economic damage associated with previous disruptions. A simple illustration of today’s diminished oil cost burden to the global economy is to adjust the nominal price of oil not only for inflation but also for the efficiency improvements. If one does, a hypothetical price of $115 per barrel today compares with an average price of $339 in 1980 in today’s dollars. By this measure, prices have plenty of runway before the oil burden resembles 1980.

But it must be balanced against other factors pulling in the opposite direction.

Oil consumption today is more concentrated in high-value uses and in areas where there is no substitute, like road or air freight and maritime shipping. These are load-bearing economic activities, less price sensitive than discretionary or consumption-oriented drivers of growth. Once disrupted they are likely to cascade through the economy... Oil concentrated in high-value uses is a little bit like rare earths, tiny compared with the size of GDP but essential for much of it. If the size of a supply disruption requires demand to come down and prices surge to the required level, the response will be sudden with a potentially unforeseen and disproportionate impact on economic activity.

9. Jakarta's success with mass transit appears exemplary.

In 2014, Jakarta was crowned the world’s most congested city by the Stop-Start Index and a year later was ranked far below other Asian cities on livability by the Economist Intelligence Unit. Ten years later, Jakarta has the world’s largest and one of the most used bus rapid transit (BRT) systems. The old, crowded diesel commuter trains, famous for allowing passengers to ride on the roofs, are now electrified, air conditioned, and run on regular schedules linking the suburbs to the city center. There are multiple subway and light rail lines crisscrossing the city. The transformation has been remarkable: in 2015, less than 20% of residents were within walking distance of transit. Now, nearly 90% of the city has access to BRT or trains... 

The user experience has also been streamlined. All of the new MRT lines are part of an integrated digital fare system. A journey from one end of Jakarta to another is capped at 10,000 rupiah (about 70 US cents). According to WRI Indonesia, as of 2024, 10 percent of trips in Greater Jakarta are now made by public transit, compared to just 2 percent in 2015... While the JICA loans covered the cost of building the MRTS and training local staff, the full cost of operation has now fallen to the city government. So far, this has worked reasonably well, with Jokowi arguing that the cost of running the system — at about 800m billion rupiah (US$50 million) a year — is justified by the estimated 65 trillion rupiah (US$3.5 billion) in annual economic losses due to traffic.

However, much more remains to be done. 

The city has recently overtaken Tokyo as the world’s largest city, with a metro population of over 41 million people, and it is still growing rapidly. It is projected to add another 10 million people in the next 25 years. To serve this population, Greater Jakarta has only six train lines and under 250 miles (400 km) of track. Tokyo, by contrast, has an astounding 158 train lines and 2,930 miles (4,715 kilometers) of track connecting 2,210 stations throughout its massive metro area.

10. Foreign portfolio investors sharply cut exposure to Indian software firms.

Foreign Institutional Investors’ (FIIs) allocation to the Indian technology sector stood at an all-time low of 7.3% at the end of March compared to 10.1% at the end of FY25, according to brokerage Motilal Oswal Financial Services... So far in 2026, FIIs have sold $21.6 billion of Indian equities – more than the $18.9 billion they sold in all of 2025 – of which $2.4 billion worth of net sales have been of Indian IT stocks (until April 15). As a result, FII holdings in the IT sector reduced to $41.4 billion from $59.8 billion at the end of 2025.

11. TIDCO stake in Titan Company.

This state government-owned industrial promotion agency, TIDCO, has a 27.88% stake in Titan Company Ltd. That’s more than the combined 25.02% share of the Tata Group through its holding company and various subsidiaries. It makes TIDCO the main promoter of Titan, which was established in 1984 as a joint venture for manufacturing quartz analog watches at Hosur on the Tamil Nadu-Karnataka border. That company — originally Titan Watches Ltd — has today grown to a premier lifestyle accessories maker with a product portfolio spanning watches and wearables (Titan, Fastrack and Sonata brands), jewellery (Tanishq and CaratLane), eyecare, women’s bags (IRTH) and ethnic wear (Taneira sarees and tops). Titan Company earned a net profit after tax of Rs 3,337 crore on a total income of Rs 60,942 crore during the year ended March 31, 2025... at the share’s closing traded price on May 6... TIDCO’s 27.88% holding in Titan will alone be valued at Rs 1,07,873 crore... if TIDCO were to sell its entire stake in Titan Company, the Tamil Nadu government would be able to mobilise upwards of Rs 1 lakh crore and bring down its outstanding debt by roughly a tenth. The annual savings in interest outgo resulting from it will far exceed the Rs 272 crore that TIDCO received as dividend for 2024-25 from its 27.88% shareholding in Titan Company.

12. Ruchir Sharma makes some very important points about the critical role being played by retail investors in the US equity markets. Some numbers.

The share of US households that own stocks has surged this decade to nearly 60 per cent, the highest proportion in any country. Americans are all in on the market, holding more wealth in stocks than in their homes for the first time. And retail is now the most active class of traders as well. Retail’s share of daily trading in US stocks doubled in the past 15 years to 36 per cent, surpassing that of big banks or hedge funds, and making them the market price-setters. Last year US retail trading topped $5tn, exceeding the pandemic high, only this time Americans weren’t stuck at home or flush with savings.

He points to contributors to the rise of retail investors.

Three forces are encouraging small investors’ deep faith in the stock market: stimulus, bailouts and technology. Record sums of money pouring out of government and central banks, intended to lift the fortunes of the real economy, have instead been used by households (particularly the richer ones) to invest in the stock market. With policymakers rushing to rescue the economy at the slightest hint of trouble, investors have come to believe the government will always bail them out. And low-cost, mobile trading platforms have given everyone easy access to investments of all kinds.

Finally, a very important political economy dynamic arises from this trend. 

The larger the retail community gets, the more pressure builds on politicians to support the market. What was said of Wall Street banks after the crisis of 2008 can now be said of the stock market as a whole: it’s “too big to fail”.

13. The rise and rise of stock market concentration in the US.

Analysts at UBS said that a measure of how many stocks were materially contributing to the index’s performance — so-called “effective constituents” — hit a record low of 42 last week, far below the level of about 100 that has been typical in recent decades.

14. Nice interactive story of how oil from the Gulf reaches Japan, gets refined, and is distributed. 

15. The $725 bn capex spending projected by the Big Tech (Amazon, Alphabet, Microsoft and Meta) is estimated to leave them with their lowest free cash flow since 2014. 

16. Fascinating 5km split of Sebastian Sawe's sub-two-hour marathon.

Some low hanging fuits in urban planning

I have blogged on multiple occasions, highlighting the importance of instruments of urban planning in shaping the form of urban development, urban renewal, economic growth, and resource mobilisation. 

This post will point to nine urban planning levers that are low-hanging fruits whose implementation can go a long way in creating the conditions for sustainable urban development. 

1. Higher FAR (without restrictive conditions) - The objective should be to incentivise more built-up area on a plot of land, especially in built-up cities. Today, a reform involving an increase in FAR is often gated by minimum plot size and minimum road width requirements, which prevent most plots from qualifying. The reform should focus on relaxing these conditions, to the extent possible, so that more plots can be built vertically, easing housing supply and using land more efficiently. A proposal in this regard is here. 

2. TDR for land foregone and a trading platform - When a landowner gives up land for public use (road, green belt, community facility, etc.), pay them not in cash but in a Transferable Development Rights (TDR) Certificate, which can be used elsewhere to build extra. This avoids land acquisition disputes, eliminates fiscal expenditures for the government, and gets roads and amenities built faster. This should be facilitated with a simple online exchange (like a stock market for TDRs) where buyers and sellers find each other transparently and in a regulated market. For each city, the numbers and values of TDRs allocated should be monitored. 

3. TDR for affordable housing / urban renewal - Use the same TDR mechanism to incentivise specific outcomes, such as affordable housing units, slum redevelopment, and heritage preservation. The critical requirement here is to be generous with the provision of the TDR to incentivise the outcome. The TDRs would be issued and transacted on a platform. The number and value of such TDRs allocated should be monitored. 

4. Purchaseable FAR and a digital transacting platform - Treat the basic FAR (say, 1.5) as a property right of the landowner, but make the remainder of the master-plan FAR (say, up to 3.5) available for purchase from the city. This generates substantial revenue for ULBs, gives developers a clear, predictable path to higher density, and removes the discretion-driven approvals that breed corruption. The purchasable FAR can, in turn, be sold at a lower rate to encourage certain outcomes, as with TDRs for affordable housing or urban renewal. The number and value of FAR purchases should be monitored. 

5. Transit-oriented development (TOD) - Allow significantly higher FAR within walking distance (typically 500m) of metro and RRTS stations. The result would be that people living and working close to transit drive less, congestion eases, and the public investment in transit gets repaid in higher tax revenue and ridership. This should be made a mandatory requirement before sanctioning of any mass transit project. The number of stations and the extent of area covered, and the incremental FAR claimed, should be monitored. 

6. Mixed-use on/beside metro stations - Build offices, shops, and homes directly above or right next to metro stations, and not several minutes' walk away in the sun. This is how Hong Kong, Tokyo, and Singapore work. In India, DDA's Karkardooma (1,800 flats) is the lone meaningful example. Every other metro/RRTS station is a missed opportunity. The numbers and descriptions of such developments should be captured and disseminated widely for emulation. 

7. Land Value Capture (LVC) - When public infrastructure (a metro line, an expressway, a flyover) raises nearby land values, capture some of that uplift through a betterment levy or impact fee, and use it to repay the infrastructure investment. Hyderabad ORR is the Indian benchmark. Today, most of the value uplift goes to private landowners. Mandate the incorporation of the LVC framework of the MoHUA in the Development Control Regulations of all cities. The city-wise numbers of LVC tools in place and the amount of revenues collected should be monitored. 

8. Business Improvement District (BID) / Tax Increment Financing (TIF) - Both are forms of LVC. In the case of BID, the businesses in a defined commercial area collectively agree to pay an extra small levy that is ring-fenced for upgrading and maintaining their own area - cleaning, lighting, security, beautification, etc. In the case of TIF, the future tax growth from a regenerated area is earmarked upfront to finance the regeneration itself. Both require enabling state legislation, and neither exists in India yet. The number of such BID/TIF, their area extents, and their respective annual realisations should be monitored. 

9. Town Planning Scheme (TPS) - It is a land readjustment mechanism used for planning urban development by pooling adjacent land parcels, developing the consolidated area with infrastructure like roads, parks, and utilities, and returning a portion (usually half) of the developed, serviced land to the original owners. This is used extensively in Gujarat. Its adoption should be encouraged in the initial stages with incentives for the development using TPS. The numbers, landowners involved, and extents of land should be monitored. 

While many states in India will claim to have issued rules adopting these, very few have implemented them in any meaningful manner to generate outcomes. Some have not been implemented at all in any city. The details matter for their effective implementation (or non-implementation). It is for this reason that the outcome metrics outlined above for each of these reforms should be monitored closely. This would force attention to the constraints that prevent their effective adoption. 

None of these is likely to bear significant outcomes immediately. But they are the plumbing for sustainable urban development and address intractable problems like affordable housing, traffic congestion, and municipal revenue augmentation. There’s a strong case for central and state governments to mandate or incentivise the effective implementation of these reforms. 

The problem of managing Chinese FDI to prevent another dependency

While most of the attention has been focused on addressing the flood of Chinese manufacturing imports, another issue that is rapidly emerging as a matter of concern is the rising Chinese FDI globally. The investment problem is rapidly emerging as an important issue, just as the trade issue. It is also an opportunity to ensure that opening up to Chinese FDI does not end up creating another dependency like that which has happened with Chinese imports. 

The investment dimension assumes great importance given the rising wall of trade barriers that restrict access to imports from China. In many countries, this includes rising measures to trace and restrict trade re-routing through third countries to avoid barriers. Further, China’s near-complete dominance in industrial metals and many other manufacturing inputs, in addition to manufactured products, means that dependence on Chinese firms is unavoidable for the foreseeable future. 

China’s sheer scale of production, especially its ability to supply metals and inputs at prices lower than almost anyone else, has pushed other producers and countries to the margins. This creates a real dilemma: if you try to restrict or control these imports, downstream industries struggle to survive due to higher costs; but if you don’t, dependence only deepens, and domestic capability erodes further. It’s a classic Catch-22, with no easy way out.

Complicating matters is the increasing strategic dimension of China’s manufacturing dominance. The former US Secretary of State Jake Sullivan has recently written that China is pursuing a theory of power that “places production, scale, and control of critical inputs at the centre of its national strategy”.

Chinese leaders are seeking to make the rest of the world dependent on China while making China independent from everyone else. And they have assessed that to achieve this, China does not need to lead in every frontier domain. Instead, it needs to control nodes of leverage—that is, the inputs and systems that advanced economies and militaries depend on to function. Beijing has already captured several of these nodes, including processed rare earths, precursor ingredients for pharmaceuticals, and batteries, and it is striving to capture others, such as robotics.

On the investment side, China is now a major outward investor globally. The graphic below shows that even as its inbound FDI has dived, China’s outbound FDI has surged. 

Debashis Basu provides some interesting numbers.

With a cumulative overseas investment stock of $3 trillion-3.5 trillion and annual outflows in the range of $160 billion-190 billion, China is a major FDI player across every region. Asia absorbs close to 70 per cent of China’s outward investment stock — roughly $2 trillion-2.2 trillion — largely into Southeast Asia, where Chinese firms have transplanted entire manufacturing ecosystems, particularly in electronics, textiles, electric vehicles, and intermediate goods. Latin America is the second major destination, with a cumulative Chinese investment estimated at $300 billion-500 billion. Europe, while smaller in share, has still absorbed between $100 billion and $200 billion. Africa, though accounting for a smaller share — perhaps $50 billion–100 billion — occupies a strategic position. West Asia, meanwhile, has attracted multi-billion-dollar annual investment in energy, petrochemicals, and, increasingly, renewables.

While the US under President Trump has gone the way of imposing strict restrictions, others, including the EU, have been grappling with wooing Chinese FDI, but with conditions. In fact, the EU has emerged as Ground Zero on this issue. Electric Vehicles and battery manufacturing are the most high-profile examples. 

Complicating matters is a collective action problem. While the EU as a whole commits to striving for reducing supply chain dependencies, individual members have an interest in benefiting from Chinese FDI and allowing their countries to be used as the source for exporting to the EU common market area. This has played into China’s hands, as it has pushed low-value-added assembly into these countries. Viktor Orban’s Hungary and Pedro Sanchez’s Spain have been the two Trojan horses for Chinese FDI into the EU. 

Another example is the UK’s decision to allow China’s Lomon Billions (LB) Group, the world’s largest producer of titanium dioxide (TDO), to buy a bankrupt TDO plant in Teesside owned by Venator Materials UK for $70 million. The Teesside plant went into administration last October with the loss of 270 jobs after more than 50 years making TDO, an industrial whitening agent used in paints and plastics as well as strategic defence and green energy supply chains. TDO is also an input to make titanium metal, a key component in the defence industry. 

The deal has faced strong criticism from European and US TDO producers who fear that the LB Group will use the plant as a source to flood the European markets with cheap TDO and drive out its manufacturers. As an illustration, the LB Group produced titanium dioxide at $1,500 a tonne in China, including subsidies, nearly half the estimated $2,800 a tonne cost to produce in the UK. 

The scenario posed by critics is real since TDO is a good illustration of how cheap Chinese exports have destroyed manufacturing capacities globally. 

LB Group produced titanium dioxide at $1,500 a tonne in China, including subsidies, nearly half the estimated $2,800 a tonne cost to produce in the UK... China became a net exporter of titanium dioxide after 2010, with exports rising from just 48,000 tonnes that year to more than 1.7mn tonnes in 2025, creating a global glut of excess production that coincided with a wave of factory closures outside China. Over those 15 years, factories with a combined capacity of nearly 1.3mn tonnes were shut down in Asia, Europe and the US, according to data compiled by industry analyst Reg Adams, who has tracked titanium dioxide markets since 1993. Chinese capacity hit 5.7mn tonnes at the end 2025.

In fact, the LB Group has made explicit its market strategy.

“By establishing new factories overseas, the company can directly connect with end-markets for production and sales, radiate out to surrounding markets . . . and circumvent high anti-dumping duties.” 

The challenge then is to manage Chinese FDI in a manner that it does not end up replacing one dependence with another, Chinese imports with Chinese firms. Instead, the objective should be to use Chinese FDI to build domestic capabilities and reduce reliance on China’s supply chain. 

This is going to be easier said than done. Chinese investments are unlikely to come cheaply. 

Since President Xi Jinping has already made it clear that China sees supply chain dependence as a strategic security lever (or “weaponised”), Chinese firms will obviously resist and stonewall efforts to use interdependence them to reduce supply chain dependence. In the circumstances, this is going to be a game of cat and mouse where the best that can be expected is a long drawn series of extractions and concessions. China’s FDI hosts should strive for two steps forward and one step backwards. 

So what are the strategies being pursued by countries to attract Chinese FDI?

The most comprehensive action plan has been outlined by the EU. The EU’s recently announced Industrial Accelerator Act (IAA) aims at increasing European manufacturing competitiveness and reducing supply-chain dependencies, and includes provisions to exercise control over foreign investment. Member states can veto any FDI exceeding €100mn in strategic sectors if the investor is from a country with more than 40 per cent of global manufacturing capacity. 

Further, these FDI projects must fill at least half of their jobs with EU workers and satisfy three of five other conditions. The five conditions are a 49% limit on foreign ownership or voting share, a joint venture structure of the investment with EU partners, licensing of IP and technology to the EU target or asset, incur R&D expenditure of at least 1% of gross annual revenue within the EU, and source at least 30% of inputs from the EU. 

An FT article nicely sums up the backdrop for the announcement of IAA.

The law is seen as the EU’s answer to decades of practices in China requiring foreign companies to invest alongside joint-venture partners and transfer their technology to local manufacturers. As Chinese manufacturers have acquired technology and been able to make higher-value products, rivals in Europe complain that they can no longer compete.

China has predictably reacted strongly to the new law and threatened countermeasures. This response is delightfully ironic in coming from a country that did the same for nearly three decades in its journey to establish manufacturing dominance. China’s commerce ministry put out a statement

…runs counter to basic market economy principles such as commercial voluntariness and fair competition… the act violated a range of WTO principles and international agreements governing intellectual property rights and subsidies… It will drag down the EU’s green transition process, damage fair competition in the EU market and bring new shocks to multilateral trade rules… (the EU should delete) “discriminatory requirements against foreign investors, local content requirements, mandatory transfer of intellectual property rights and technology requirements, public procurement restrictive policies, and other content.

There’s a deep irony in these measures coming from Europe, and they also represent the wheel turning the full circle. In their early years of industrialisation, the US and Europe pursued infant industry protection policies to build domestic manufacturing capabilities. Once they became dominant, they embraced the principles of free trade and forced developing countries to lower tariffs and open up their markets. Now, having given up their manufacturing dominance to China, the same countries are back to pursuing protectionist policies to safeguard their industries from Chinese competition. 

Another example of Chinese investments that should be watched with interest is that made in Indonesia’s nickel processing and refining, and now being made in the same country’s alumina processing and aluminium manufacturing.

In 2014, Indonesia banned the export of raw Nickel, with a total ban from 2020, thereby encouraging Chinese companies, led by the steel giant Tsingshan Holding Group, to set up processing plants across the country. The transformation since has been remarkable for a country that, a decade ago, was not even a major nickel producer.

Although Indonesia held the world’s largest reserves… most of it was low-grade nickel that it had not yet figured out how to process efficiently… Before 2014, most of Indonesia’s nickel ore was sold to nickel and steel manufacturing plants in China… Huge investments came from Chinese steel, nickel and battery manufacturers, including Tsingshan, CATL and Lygend, who partnered with Indonesian mining companies to set up processing facilities. Chinese stakeholders control over 75 per cent of Indonesia’s refining capacity, according to a recent report by C4ADS, a Washington-based security non profit. Not only did the companies bring capital, they also brought the knowledge to process Indonesia’s low-grade nickel reserves quickly and profitably… The Chinese had made advancements in rotary kiln electric furnaces, which turns nickel ore into raw material for steel. They had also mastered the high-pressure acid leach technology, a refining process that converts low grade nickel ore to battery-grade — a procedure that western companies had struggled with for years…

Indonesia has gained control of the market and cemented itself as the epicentre of global nickel production for years to come. Last year, Indonesia accounted for 61 per cent of the global refined nickel supply up from just 6 per cent in 2015… Its market share is expected to grow to 74 per cent by 2028. This means Indonesia now controls more of the world’s supply of nickel than Opec did of oil at the cartel’s peak in the 1970s — then around half of global crude oil output. 

And it has achieved this position of market dominance just at the moment when new customers are desperately trying to secure reliable supplies. Not only are carmakers like Tesla, Ford and Volkswagen racing to source the metal for lithium-ion batteries, nickel is also widely used in smartphones and other electronics — as well as being a vital ingredient in stainless steel… Surging production in Indonesia has wiped out competition from companies such as Australian mining group BHP and dramatically reshaped the global supply chain. By flooding the market and driving down prices, the Sino-Indonesian partnership has made it much harder for rivals to produce the metal economically elsewhere in the world.

Similarly, in 2023, Indonesia banned the export of bauxite, the mineral used to make Aluminium, forcing companies to establish domestic refining (bauxite to alumina) and smelting (alumina to aluminium) facilities. Chinese processors, already facing restrictions on capacity expansion in China, moved to Indonesia to invest in the high-margin industry. A combination of Chinese production caps, sanctions on Russian smelters, closures in Africa and Australia due to rising energy costs, and now the damage to facilities in the Gulf has squeezed supply and is driving up prices. This, in turn, could accelerate Indonesia’s smelters build out, especially by the Chinese companies. The country produced 5.9mn tonnes of alumina last year, up from 3.3mn in 2022. 

The Indonesian experiment with Nickel and Aluminium export bans and forced domestic processing, primarily with Chinese investments, is a test case for leveraging Chinese companies and their technologies. The extent of localisation and domestic value addition, and technology transfer and spill-overs that have happened in these two industries is not known. Clearly, given the strategic nature of these metals and the attendant leverage, the Indonesians may have missed a trick by not insisting on joint ventures with domestic firms.

Infrastructure is sometimes considered another destination for Chinese FDI. However, as I have blogged here, China is unlikely to undertake FDI in infrastructure. Instead, its primary channel for funding infrastructure is through loans to governments, as has been institutionalised through the Belt and Road Initiative (BRI). This channel is filled with pitfalls and problems that are well documented. 

Even equity investments in infrastructure come with strategic risks. An illustration of this is that in ports.

Chinese firms now operate or have a financial stake in at least 129 ports outside China, and have spent at least $80bn on port construction from Antigua to Tanzania, with many of the investments tied to bilateral trade and regional shipping agreements. More than a third of China’s overseas ports are near maritime chokepoints, including the Strait of Malacca, the Strait of Hormuz and the Suez Canal, making them indispensable operators in strategic areas. China’s firm grip on global ports has rattled Western governments. MERICS, a think-tank in Berlin, found that after a terminal operating contract is signed, total trade with China rises by more than a fifth, while countries that allow Chinese firms to run all their terminals at one of their ports see a 19% drop in exports to the rest of the world. Operating ports allowed Chinese firms to prioritise their cargo and vessels and speed up customs and logistics...

China’s reach extends beyond physical infrastructure. LOGINK, a Chinese government-run logistics-management software, is used in at least 24 countries and 86 ports (America banned its use in 2023). LOGINK shares data with CargoSmart, another shipping-management software firm owned by COSCO, another Chinese state-owned firm, and in turn gives it access to the whereabouts of 90% of the world’s container ships. It also has a tie-up with CaiNiao, a logistics provider with hundreds of warehouses around the world... Chinese firms are also building industrial parks and manufacturing facilities close to their existing ports in Africa and Europe.

So what does all this mean for India?

Debashis Basu again provides these numbers on FDI from China.

Chinese FDI in India amounts to a trivial $2.51 billion, or 0.32 per cent of India’s cumulative equity inflows, since 2000. Even if one broadens the definition to include venture-capital investment and indirect flows routed through third countries, the total rises only to $15 billion-20 billion.

The Economic Survey 2023-24 also proposed that India should seek to attract Chinese FDI to benefit from the ‘China plus one’ strategy embraced by foreign multinationals seeking to diversify away from that country. 

The challenge, like with all things in public policy, is about how to do it. For a start, it might be useful to study the EU’s IAA. It is important to prepare a multi-track strategy that places different conditions for sectors depending on their importance and the extent of Chinese dominance. I had blogged here outlining some mitigation strategies for India as it grapples with chokepoints in trade. 

It must be a conscious strategy of attracting Chinese investors with some broad requirements and then gradually tightening the conditions to first increase domestic value addition and then gradually technology transfer to domestic suppliers. This will be a game of hardball, and India must be prepared to play it with China. 

It has had opportunities to pursue this strategy earlier. In 2020, one analysis reported that 16 out of India’s then 29 unicorns had Chinese investors, with them being lead investors in eight. Tencent and Alibaba were the largest investors. As relations soured, these investors were forced out or left.

Basu’s description of how China itself courted foreign investment is appealing in theory, but very hard to execute for several reasons, especially given the current circumstances of countries having to deal with China.

Beginning in the 1990s and accelerating through the 2000s, China actively courted FDI and technology, particularly from Japan, the United States, and Europe, in a period known as “reform and opening up” (gaige kaifang). China did not see foreign capital as a threat; it saw it as an instrument of transformation if the state retained strategic control over direction. Multinationals were often required to form joint ventures, localise production, and, in many cases, transfer technology — the phrase “bring it in” (yin jin lai) capturing this mindset. To take full advantage of this, special economic zones provided infrastructure, policy stability, and export incentives, enabling China to integrate into global manufacturing networks and develop deep supplier ecosystems, skilled labour pools, and process expertise. China used this phase to absorb technology, build domestic champions, and gradually move up the value chain, eventually becoming both an exporter and an investor in the next phase — “go out” (zou chu qu).

I’m not sure that Chinese manufacturing firms will come in with low minority stakes just because of the Indian market size. Historically, it has not happened, even in much larger and more attractive markets like the US and Europe. And Chinese FDI has generally steered clear of India. 

This strategy will require high institutional capabilities that track emerging FDI trends and global developments, and dynamically shape and steer India’s FDI policy, especially with respect to China (or countries with land borders). This will have to be quite a different approach from that which has generally been followed by India.