Narendra Modi’s state visit to Washington this week produced all the symbolism expected from one of America’s highest forms of diplomatic recognition. President Joe Biden hosted him at the White House, Modi addressed a joint meeting of Congress for the second time, and both governments repeatedly described India and the United States as increasingly important strategic partners. Yet the more consequential parts of the visit were considerably less ceremonial. The two countries announced or advanced cooperation involving semiconductor manufacturing, jet-engine technology, drones, space, telecommunications, artificial intelligence and other strategically important industries. (whitehouse.gov)
Last week, I argued that India is not becoming an American ally in the traditional sense. Nothing this week changes that conclusion. India has not abandoned strategic autonomy, joined a US-led treaty alliance or adopted Washington’s position on Russia. What has changed is the economic content of the relationship. The United States increasingly appears willing to help India acquire capabilities that Washington historically guarded much more closely, while India is attempting to use America’s desire for alternatives to China to accelerate its own technological and industrial development.
The proposed GE Aerospace-Hindustan Aeronautics arrangement to manufacture GE F414 fighter jet engines in India illustrates the change particularly clearly. India has purchased foreign military equipment for decades, but purchasing a finished engine and acquiring the technology and manufacturing capability required to produce sophisticated engines domestically are very different forms of economic relationship. The first creates military capability. The second can also create industrial capability, engineering knowledge and potentially spillovers into other sectors. (whitehouse.gov)
This is why technology transfer matters so much to developing powers. Countries rarely become technologically advanced merely by importing increasingly sophisticated finished goods. They need firms, engineers and institutions capable of understanding, adapting and eventually producing complex technologies themselves. South Korea, Japan and China followed very different development models, but all eventually moved from importing substantial foreign technology towards developing domestic technological capabilities.
India has attempted this repeatedly, with mixed results. Its space and nuclear programmes demonstrate that Indian institutions can develop sophisticated indigenous technology when political commitment, technical talent and long-term investment align. Manufacturing performance across the wider economy has been much less consistent. The challenge is therefore not proving Indian engineers can master advanced technology. It is creating industrial ecosystems capable of reproducing that success at scale.
Semiconductors may be an even larger test. US company Micron announced plans for a semiconductor assembly and testing facility in Gujarat, supported by substantial incentives from the Indian government and Gujarat. (micron.com) The project is significant, but it is important to describe what it actually represents. Assembly and testing are valuable parts of the semiconductor supply chain, but they are not equivalent to manufacturing the world’s most advanced logic chips. India is entering a highly complex industrial ecosystem gradually rather than suddenly becoming another Taiwan.
That distinction matters because governments frequently announce investment totals as though money automatically produces industrial capability. A semiconductor ecosystem requires reliable electricity, ultrapure water, specialised chemicals, logistics, skilled engineers, suppliers and extremely demanding quality control. Subsidising one factory can attract investment, but the larger economic return arrives when suppliers, skills and complementary industries develop around it.
This is the difference between a factory and an ecosystem. A factory produces output. An ecosystem develops the capacity to create additional factories, suppliers and technologies without requiring the government to reconstruct the entire system for every investment.
India’s industrial objective should therefore be larger than attracting multinational corporations through subsidies. Foreign investment becomes developmentally transformative when domestic firms and workers learn from it, supply local inputs, move into higher-value activities and eventually generate technologies of their own. If imported components arrive, foreign companies assemble them inside subsidised Indian facilities and most sophisticated knowledge remains elsewhere, employment may increase without producing the deeper technological transformation policymakers want.
Economists describe this through spillovers. Foreign direct investment can transfer knowledge indirectly as employees move between firms, suppliers learn higher production standards and domestic companies imitate new processes. These spillovers are not automatic. They depend on local firms possessing enough absorptive capacity to learn from foreign investors.
Education therefore returns to the centre of industrial policy. India’s demographic advantage matters only if its workforce can supply technicians, engineers, researchers and managers at the scale advanced manufacturing requires. Industrial strategy cannot be separated from universities, vocational education and school quality. A semiconductor subsidy announced today depends partly on mathematics being taught effectively in classrooms years earlier.
This is where the US-India partnership connects directly to the population argument from April. India’s young population creates an enormous potential workforce, while geopolitical diversification creates demand for production outside China. Those two trends create an unusual opportunity. But neither guarantees India captures the investment. Vietnam, Mexico, Indonesia and other countries also want supply chains relocating or diversifying from China.
India therefore competes not merely through lower wages but through institutional reliability. Investors care about infrastructure, taxation, customs procedures, contract enforcement, regulation and whether policy remains predictable over decades. A semiconductor facility can cost billions and operate for many years. Companies therefore need confidence that the rules surrounding today’s investment will remain sufficiently stable tomorrow.
This is why industrial policy is ultimately a state-capacity problem. Governments can announce subsidies relatively easily. Building institutions that allocate subsidies efficiently, prevent corruption, coordinate infrastructure and withdraw support from failing projects is much harder.
Industrial policy fails when governments become incapable of distinguishing strategic investment from politically connected rent-seeking. Once firms discover that lobbying produces higher returns than innovation, public subsidies stop creating capability and start creating dependency.
The successful East Asian examples often involved substantial government intervention, but intervention was frequently connected to performance requirements. Firms received support while being expected to export, increase productivity or meet technological objectives. The state did not simply choose companies and protect them indefinitely from competition.
India needs a similar principle: support should purchase capability, not permanence.
This applies particularly to Modi’s broader “Make in India” strategy. Domestic manufacturing is valuable because it can create employment, technological learning and strategic resilience. But producing something domestically is not automatically economically superior to importing it. If Indian consumers permanently pay twice the global price for an inefficient domestic product protected from competition, national self-sufficiency can become an expensive political slogan.
Strategic sectors justify greater costs when dependence itself creates substantial risk. Jet engines, semiconductors, telecommunications and certain defence technologies plausibly meet that threshold because access could become constrained during geopolitical conflict. Ordinary products require a much stronger economic case before national security becomes a justification for protection.
The G7’s “de-risking, not decoupling” principle therefore applies to India too. New Delhi wants to reduce strategic dependence on China without eliminating trade with China. China remains one of India’s largest trading partners, and Indian industries rely on Chinese inputs across several sectors. Attempting rapid economic separation would impose substantial costs.
The objective should be reducing critical dependencies, not trade itself.
This distinction is especially important because geopolitical competition can tempt governments into confusing resilience with autarky. A resilient India should be capable of obtaining important technologies from several sources and producing some domestically. An autarkic India would attempt to produce everything itself regardless of comparative advantage.
The first increases strategic options.
The second can reduce prosperity.
The United States faces its own version of this calculation. Washington wants stronger Indian capabilities partly because a more powerful India changes the Asian balance of power. Helping India manufacture jet engines or develop semiconductor capacity therefore serves American strategic interests even if it also creates a more technologically independent India.
That is an interesting departure from traditional patron-client relationships. The United States is not simply trying to make India dependent on American security guarantees. In several areas, it is helping India become more capable of acting independently.
This can work because Washington increasingly believes Indian autonomy is preferable to Chinese dominance.
The distinction reveals how multipolarity changes great-power strategy. A hegemonic power might prefer weaker partners that remain dependent on it. A power facing a serious peer competitor can benefit from partners becoming stronger even when those partners refuse complete alignment.
India’s value to Washington therefore comes partly from the fact that India is not an American satellite. A country of 1.4 billion people with its own military, nuclear arsenal, economy and diplomatic network can shape Asia independently in ways that often constrain China without requiring direct American control.
Strategic autonomy becomes an American asset when Indian and American interests overlap.
The limits become visible when they do not.
Russia remains the obvious example. India continues purchasing Russian oil and maintains substantial defence relationships with Moscow despite American efforts to isolate Russia after the invasion of Ukraine. Washington has increasingly tolerated those differences because confrontation with India could damage cooperation on issues it considers more important over the long term.
That is hierarchical prioritisation of interests. Governments do not need agreement on everything. They need to decide which disagreements are important enough to block cooperation elsewhere.
India is doing exactly the same thing. It disagrees with American approaches to Russia, trade and aspects of global governance while concluding that cooperation on technology, China and investment remains valuable.
This is what a mature strategic relationship should probably look like. Constant agreement can indicate either extraordinary alignment or dependency. Durable partnerships often contain structured disagreement.
The democratic-values issue remains more uncomfortable. Modi’s visit generated criticism from some American lawmakers and human-rights organisations concerned about religious freedom, press freedom and democratic institutions in India. Biden nevertheless offered Modi the full ceremony of a state visit.
This exposes the selectivity inherent in values-based foreign policy. The United States describes democracy as an important element of its global strategy but inevitably gives greater strategic consideration to the domestic politics of powerful partners than weaker states.
That inconsistency should be acknowledged rather than disguised. Washington can reasonably conclude that strategic cooperation with India is important while still raising concerns about democratic institutions. What weakens credibility is pretending the trade-off does not exist.
India should resist the opposite simplification. Criticism from American politicians or journalists is not automatically evidence of hostility towards India. A democracy claiming global legitimacy partly through its democratic character should expect its institutions to be scrutinised.
Sovereignty means India decides its own politics.
It does not mean foreign observers lose the right to have opinions about them.
The relationship will be stronger if it can survive that disagreement rather than requiring either side to perform constant ideological harmony.
The defence agreements announced this week may ultimately test that maturity. India wants access to advanced American technology without becoming strategically dependent on American political approval. The United States wants to share enough technology to strengthen India without losing control over capabilities it considers sensitive.
Both sides therefore negotiate over trust and autonomy simultaneously.
Technology transfer itself can become bargaining power. A supplier able to withhold spare parts, software updates or components retains leverage even after selling the original system. India understands this from decades of defence procurement and therefore emphasises domestic production.
Manufacturing the system locally reduces some dependence but does not necessarily eliminate it if critical intellectual property or components remain foreign. True technological autonomy exists on a spectrum.
This is why measuring “indigenisation” through the percentage of a product manufactured domestically can be misleading. A system may contain mostly locally produced components while depending on one imported component that cannot be replaced. Strategic vulnerability is determined by the least substitutable input, not simply the average domestic content.
Again, supply chains behave like networks.
The weakest critical node matters disproportionately.
The broader industrial strategy should therefore identify which capabilities would be genuinely difficult to replace during crisis and concentrate scarce public resources there. Trying to subsidise every stage of every industry would spread investment too thinly and create opportunities for rent-seeking.
India’s greatest advantage may ultimately be that geopolitical competition gives it choice. American companies want alternatives to China. European governments want diversified supply chains. Japan wants a stronger Indian economic partner. Russia wants to preserve defence and energy relationships. Gulf states want Indian investment and trade. India can negotiate among all of them.
Choice increases bargaining power because no single partner becomes indispensable.
But bargaining power is valuable only if it produces domestic capability rather than better diplomatic photographs.
The state visit will be remembered through Modi’s congressional address, the White House ceremony and the language of partnership. The economically important question is what Indian factories, laboratories and workers look like ten years from now.
Did semiconductor investment create a larger electronics ecosystem?
Did jet-engine cooperation create transferable engineering capability?
Did defence production generate domestic suppliers?
Did universities produce enough skilled workers?
Did India become more technologically autonomous, or merely replace one set of foreign dependencies with another?
Those outcomes will determine whether the visit mattered.
This is why I would not describe the emerging US-India relationship primarily as an alliance. Alliances concern who stands beside whom during conflict. What is being constructed now is broader and perhaps more consequential: a network of technology, capital, defence and supply chains intended to alter the distribution of productive capability across Asia.
China’s rise provides much of the strategic incentive.
India’s development provides much of the economic opportunity.
The United States provides technology and capital India wants.
India provides scale and strategic weight Washington wants.
None of this requires either country to surrender independent foreign policy.
That is what makes the relationship interesting.
Last week, I argued that India and America are getting closer without India becoming an American ally. This week’s agreements show what that can mean materially. Strategic autonomy is no longer simply India’s ability to say no to Washington, Moscow or Beijing.
The more important form of autonomy is having enough domestic capability that saying no is economically and technologically possible.
Diplomacy creates options.
Industrial capacity determines whether a country can use them.
