The Compartmentalised Doctrine: A Method for Middle Powers in a Two-Camp World
Visas, flights, water data, pilgrimages, and an unbroken multilateral seat held even through a cross-border military strike. India's quiet rewriting of how to deal with a hostile great power may be the most useful diplomatic export of the decade.
The great-power competition framework that has dominated global affairs for a decade is built on an implicit assumption: that countries must choose. That bilateral relationships flow from camp alignment. That engagement on any track is a concession on every track. That the worst track in a relationship sets the temperature for the rest.
This assumption has held for the United States and China, who have run a structurally bipolar relationship in which trade tension, technology competition, military signalling, and human-rights friction reinforce each other rather than being negotiated separately. It has held for Russia and the European Union, where the post-2022 freeze has been near-total. It has held, with variations, for most countries that have tried to operate one major-power relationship at a time.
But for the dozens of middle powers who must run several difficult major-power relationships simultaneously, the all-or-nothing model is not workable. They have been quietly experimenting with an alternative. The most coherent example of the alternative — and the one most worth studying — is the doctrine that has emerged in India's relationship with China since 2024.
It is best named in one word: compartmentalised.
What compartmentalisation looks like
The arc of India's China policy over the last eighteen months tells the story.
The 2024 BRICS summit produced a Modi-Xi meeting that ended five years of post-Galwan deep freeze. Patrolling protocols on the Line of Actual Control resumed at Depsang and Demchok. By 2025, India had restored regular tourist visas for Chinese nationals after a four-year suspension. Direct flights between Indian and Chinese cities resumed. The Kailash-Mansarovar Yatra reopened for Indian pilgrims. Trans-boundary river data sharing — the hydrological information that affects the lives of tens of millions in eastern India each monsoon — resumed. The Special Representatives mechanism for boundary talks met for its first substantive round in five years.
Then came two stress tests. The Pahalgam terror attack in April 2026, in which 26 Indian civilians were killed. The Indian military response — Operation Sindoor, on 7 May 2025 — which struck nine launchpads inside Pakistan and Pakistan-occupied territory. China's diplomatic posture during and after the strikes was openly supportive of Pakistan, with Chinese-supplied air defence and electronic-warfare equipment now central to Pakistan's military picture.
By any precedent in the prior twenty years of India-China relations, the deep freeze should have returned. It did not.
Visas were not suspended. Flights were not paused. The Kailash yatra schedule for 2026 was not pulled. Water-data sharing continued. The Indian delegation attended the next Shanghai Cooperation Organisation ministerial without theatre. The Special Representatives talks continued on schedule. The negotiating posture inside BRICS, where Chinese and Russian preferences increasingly diverge from Indian ones, hardened on substance but did not become performative.
This was a deliberate choice. It is the visible expression of a doctrine.
The doctrine, named
A senior Indian official, speaking off-the-record at a Delhi seminar, used the word that fits. "Compartmentalised." India and China, he argued, will be neighbours, rivals, and trading partners simultaneously for the rest of his career and the rest of his children's careers. The country cannot afford to run a single-track policy that subordinates everything to whichever track happens to be hottest in a given quarter.
The doctrine, in one sentence: do not let the worst track set the temperature for the others.
In practice it operates as four parallel tracks running at four different temperatures.
On hard security, India will hold position at the boundary, build infrastructure faster than China expects, and respond to incursions kinetically when required. That track gets a hard answer.
On trade, India will continue to allow Chinese imports of capital goods that domestic industry genuinely needs, while restricting consumer-electronics flows that compete with strategic manufacturing programmes. That track gets a calibrated answer.
On the multilateral file — the SCO, BRICS, climate negotiations, WTO posture — India will engage as a full member, contest Chinese positions on substance, and refuse to walk out of rooms even when walking out would be cathartic. That track gets a patient answer.
On people-to-people — students, pilgrims, tourists, business travellers, water-data engineers, river-monitoring scientists — India will keep the channels open, because closing them hurts Indians more than it hurts the Chinese state. That track gets a generous answer.
Why this matters beyond India
The compartmentalised doctrine is not Indian by necessity. It is Indian because India has had to run it, and because the country's diplomatic apparatus has the breadth, the institutional memory, and the bureaucratic patience to sustain a policy with four temperatures simultaneously.
But the structural problem the doctrine solves is not unique. It applies to every middle power that finds itself running difficult relationships with both China and the United States, or with Russia and the European Union, or with multiple regional rivals at once.
Indonesia is currently navigating a parallel set of pressures with both Beijing and Washington. Vietnam runs an even more delicate balance. Brazil has been forced to compartmentalise its relationship with both China — its largest trading partner — and the United States. South Africa has had to do the same with the BRICS architecture and Western capital markets. Saudi Arabia is testing it across the US-China-Russia triangle. Turkey runs a version of it across NATO membership and Russia ties.
Each of these countries faces the same underlying problem: the all-or-nothing posture of camp alignment is a luxury available only to economies large enough or distant enough that they need not negotiate trade-offs in real time.
For everyone else, the alternative model is the compartmentalised one. India's contribution is not to have invented it but to have run the most public test of it through a high-stakes military episode and demonstrated that the policy holds.
Is compartmentalisation strategic maturity or fence-sitting?
This is the central critique the doctrine attracts, and it deserves a serious answer.
The fence-sitting case is straightforward. By keeping engagement open across so many tracks during periods when the rival great power has been visibly hostile, the middle power is providing diplomatic legitimacy and economic interconnection that the rival would otherwise have to earn. Critics argue this is, at the margin, a concession that the rival prices into its own behaviour.
The strategic-maturity case is also straightforward. Severing engagement on the easy tracks costs the middle power access to information, leverage, and ordinary commercial benefit, while delivering no actual punishment to the rival's system. The post-Galwan freeze was tested by India and found wanting — it ceded ground without producing a softer Chinese position. Compartmentalisation is the policy that allows the country to be tough where toughness matters and patient where patience matters.
Both cases have weight. The doctrine will be judged on outcomes that take a decade to read, and two specific outcomes will eventually deliver the verdict.
The first is the actual hard-power balance at the contested boundary. If infrastructure, force posture, and surveillance density at the line continue to improve, compartmentalisation is the policy of a country that is using the open tracks to buy time for the hard build. If those metrics slip, compartmentalisation will reveal itself as the softer policy its critics claim.
The second is the regional architecture. If the middle power can use its multilateral seats and bilateral engagement with smaller states to peel some of those states away from automatic great-power gravity, compartmentalisation is producing leverage. If the regional architecture continues to drift toward the rival's gravity, the leverage was theoretical.
The underwriting condition
There is one underwriting condition without which compartmentalisation does not work.
The middle power has to keep growing fast enough, building fast enough, and modernising its capabilities fast enough that the asymmetry between itself and the rival narrows over time rather than widens. If that asymmetry widens, compartmentalisation becomes a polite name for accepting permanent inferiority.
This is the real conversation, and it is mostly a domestic one rather than a diplomatic one. The diplomacy is the easy part. The build is the hard part.
Compartmentalisation buys time for the build. It does not substitute for it.
Why this is a public good
A doctrine that allows multiple middle powers to run difficult major-power relationships simultaneously without being forced into formal camp alignment is a global public good. It preserves the multilateral system. It reduces the risk that local rivalries are amplified into bipolar showdowns. It creates space for the kind of issue-by-issue coalitions that the climate transition, the semiconductor build-out, and the AI-governance question all require.
The Global Federation has long argued that the bilateral camp structure of mid-twentieth-century geopolitics is not the model the rest of this century should be built on. The compartmentalised doctrine is not the only alternative, but it is the most thoroughly tested one in the field.
Watching how it travels — to Indonesia, to Brazil, to Turkey, to South Africa, to Saudi Arabia — will be one of the more useful diagnostic exercises of the next five years.
The Global Federation covers global affairs with the conviction that the most durable doctrines are usually the ones developed quietly under stress, and the most useful policies are usually those that are easy to imitate.