The year is 2036. Bangladesh joined the Mecca Joint Defence Agreement ten years ago, despite opposition at home.
A meeting is under way in New Delhi. Not the televised kind, with flags behind everyone and prepared statements in front. The other kind. The kind where a border can move on a map before the public knows there was a meeting, the kind that's for history books or takes decades to be declassified.
The service chiefs are there, along with national security and intelligence officials, senior ministers and political advisers. The diplomats have taken up a good part of the table too. Once soldiers cross an international border, the foreign ministry gets busy.
Relations with Bangladesh have been deteriorating for years. A hawkish government came to power in India after two more election cycles, and it now sees Bangladesh as a growing threat in what it considers its own backyard.
Chinese state-owned enterprises have spent the last decade building airports, power projects, communications infrastructure, logistics facilities and other projects across Bangladesh. Most of it may be commercial. Some of it may have military uses in a crisis. Nobody in that room wants to base national security policy on the assumption that the Chinese built everything for the love of civil engineering.
Bangladesh has also deepened military cooperation with Türkiye, Pakistan and Saudi Arabia. There have been exercises, arms purchases, intelligence agreements and joint planning under the Mecca Pact.
None of this proves that Bangladesh intends to attack India. But intentions can change after an election. Capabilities, once acquired, tend to remain on the balance sheet.
So an idea is placed on the table.
India could seize and hold a narrow strip of northern Bangladesh, neutralise some of the infrastructure it considers dual-use, and force Dhaka to accept limits on future Chinese involvement.
Not a full invasion. Nothing so dramatic. A limited operation. A buffer zone, perhaps. A security corridor. A temporary territorial adjustment.
Nobody says annexation at the beginning. Large military operations usually arrive under softer names anyway.
The benefits are easy to list. Control over useful positions. More distance between sensitive Indian territory and Chinese-built infrastructure. And, more importantly, a message to Dhaka, Beijing and anyone else watching: there are limits to what India will tolerate in its neighbourhood.
Now comes the less enjoyable part. What will it cost?
The calculation looks something like this:
E(C_attack) = C_baseline + Σ_i P_i × ΔC_i
The baseline is the cost of fighting Bangladesh, occupying hostile territory and absorbing the immediate economic and diplomatic fallout. The second part adds the probability-weighted cost of each possible allied response.
The responses could range from no response to diplomatic support, intelligence and weapons, financial backing, a limited military deployment or direct entry into the war. Each branch has a different probability and a different bill attached to it.
This is not an equation where someone enters five numbers and gets the answer at the bottom. It is a way of forcing everyone in the room to count the outcomes that do not appear on the battlefield map.
Ten years ago, the baseline was already unpleasant. Bangladesh is densely populated, difficult to occupy and not exactly empty of armed forces. Any operation would carry casualties, displacement, international condemnation and economic disruption.
But the equation is worse now.
Bangladesh did not join the pact and then close down its military on the assumption that Türkiye would take care of everything. It continued spending on air defence, drones, surveillance, hardened infrastructure, ammunition, fuel reserves and the ability to keep fighting after the first round of strikes.
The pact is not the first line of defence. Bangladesh still pays the deductible.
Then comes the difficult part: the probability of allied involvement.
Nobody in the room believes Turkish divisions are certain to land in Rangpur the following morning. But that is not the question.
The question is whether everyone can safely assume that Türkiye, Pakistan and Saudi Arabia will do nothing.
Türkiye could provide satellite imagery, drones, electronic-warfare support, air-defence systems or personnel. Pakistan could share intelligence and targeting information, accelerate ammunition deliveries or raise military pressure elsewhere, forcing India to divide its attention. Saudi Arabia could finance emergency procurement, apply diplomatic pressure or use its wider economic relationships.
Maybe the assistance would be limited. Maybe it would arrive late. Maybe the three countries would hold a meeting, issue a statement and provide less than Bangladesh expected.
All of that is possible. But zero is no longer the obvious number.
Now suppose a small multinational contingent is already in Bangladesh for an exercise: a Turkish air-defence team, Pakistani intelligence and logistics officers, and a smaller Saudi support unit.
A few hundred personnel cannot stop the Indian military. That is not why they are there.
Before the operation, the question in Ankara is: Should Türkiye enter a war involving Bangladesh? After Turkish personnel are killed in an attack, the question becomes:
What should Türkiye do after its own soldiers have been killed while serving under a treaty mission?
The deployment has not changed the military balance much. It has changed the sequence of decisions. Türkiye does not care about Rangpur the way Bangladesh does. Saudi Arabia probably cares even less. But they have another problem.
The Mecca Pact has existed for a decade. Its members have held exercises, signed contracts, created committees and spent years describing an attack on one as an attack on all. If a clear attack on a member produces no meaningful response, the pact has defaulted on its visible claim.
Just one default can reprice the entire guarantee.
Pakistan will start asking what the guarantee is worth for Pakistan. Saudi Arabia will ask the same. Smaller members may begin looking for separate weapons suppliers, bilateral arrangements or accommodation with the powers the pact was meant to deter.
Potential adversaries will update too. If the guarantee failed once, perhaps the next member is also worth testing. The agreement could remain alive on paper while losing most of its value in practice.
The question for Türkiye is therefore not simply whether northern Bangladesh is worth the immediate cost. It is what a Turkish commitment is worth tomorrow if Türkiye abandons Bangladesh today.
Now the economists get involved.
What happens to shipping insurance, markets, investment and the rupee? What does China do if Indian forces strike airports and infrastructure built by Chinese state companies? What happens to India’s trade, energy and investment relationships with the Gulf when Saudi Arabia is sitting on the other side of the treaty? What export restrictions might follow? How long can the operation be sustained if it does not end on schedule?
By this point, nobody can give the room what it wants. The generals cannot guarantee that the operation remains limited. Intelligence cannot assign a reliable probability to allied intervention. The finance ministry cannot put a ceiling on the bill.
India may still be capable of defeating Bangladesh in a direct conventional fight. But that is no longer the full decision.
The decision is whether a limited operation against Bangladesh remains a limited operation against Bangladesh.
Perhaps the chance of serious allied involvement is only 10 percent. Perhaps it is 20. But 10 percent is not a reassuring number when the 10-percent outcome is a regional war.
A professional military planner does not calculate only the most likely result. He also has to explain what happens in the ugly one. Some would call that a fat-tail risk.
The proposal is not rejected. Governments rarely reject ideas so neatly. It is sent back for further assessment of escalation, allied reactions, economic consequences and diplomatic fallout.
In bureaucratic language, the operation is deferred. Nothing happens in northern Bangladesh. No foreign troops arrive to save it. No missiles are fired. No emergency summit is called. The pact is never invoked.
But there is also no invasion.
No one counts the villages that were not shelled, the factories that were not closed, the families that were not displaced or the LNG cargoes that did not have to be diverted during a war.
Ten years after joining, someone in Dhaka may point out that the Mecca Pact has never been used. Yet that may have been the point all along.
This is the part missed when the value of a defence pact is measured by counting how often its final clause has been invoked.
NATO’s famous Article 5 has been applied only once, after 9/11. This is often presented as evidence that collective-defence clauses are mostly decorative.
There is another possible explanation for the shortage of Article 5 cases: countries have shown a remarkable reluctance to invade NATO members.
A high Article 5 invocation rate would not necessarily constitute excellent customer satisfaction.
After 9/11, the allies provided intelligence, security at allied facilities, overflight clearances, access to ports and airfields, and NATO AWACS patrols over the United States. The United States did not need Latvian infantry to defend Washington. It requested particular forms of assistance that the allies could provide.
There is a large space between doing nothing and Pakistan dropping a nuclear bomb on India during the hypothetical war-room scenario above. Much of defence policy takes place there.
Article 5 itself does not require every ally to declare the same war, in the same way, at the same time. It requires each member to take “such action as it deems necessary, including the use of armed force.”
That could mean combat forces. It could also mean intelligence, logistics, surveillance, air defence, maritime support, weapons, financing or access to military facilities.
Article 4 is not as popular among barstool diplomats as Article 5. A consultation followed by a radar deployment is less cinematic than a nuclear exchange, which may explain the difference.
Article 4 allows any NATO member to request consultations when it believes its territorial integrity, political independence or security is threatened. It has been invoked by Türkiye, Poland, Estonia and groups of eastern members, not merely by the United States or the larger European powers.
Türkiye invoked Article 4 in February 2003 because of the possible consequences of war in neighbouring Iraq. NATO subsequently deployed AWACS surveillance aircraft, Patriot air-defence batteries and chemical, biological, radiological and nuclear defence support. NATO, as an organisation, did not join the US-led invasion of Iraq.
That distinction matters. Türkiye did not have to wait for an Iraqi missile to land before the alliance became useful.
The response to Crimea followed the same logic. After Russia annexed Crimea in 2014, NATO expanded air policing, surveillance, exercises and readiness measures across its eastern flank. The multinational battlegroups in Estonia, Latvia, Lithuania and Poland were approved in 2016 and deployed in 2017.
These were not armies large enough to defeat Russia by themselves. They were tied to a much larger system of planning, reinforcement and command. Anyone attacking them would not be opening a clean bilateral war against Estonia or Latvia.
Russia has invaded non-NATO Georgia and Ukraine. It has not attempted a comparable conventional seizure of Estonia, Latvia or Lithuania. NATO may not be the only reason. But the comparison is difficult to ignore.
Why would France or Britain help defend NATO ally Latvia? Neither country has a deep emotional attachment to Latvian territory.
Because alliances do not operate as one-off acts of kindness. They are sequential games.
France is not considering only Latvia. It is considering what abandoning Latvia does to the next round. Russia updates its belief about NATO’s willingness to fight. Lithuania and Estonia update their beliefs about the guarantee. Poland begins thinking about alternatives. If France later expects other members to treat a threat against France as a common problem, they will remember whether France treated the Latvian claim as an obligation or an optional favour.
France cannot expect the alliance to operate as a one-way facility, compulsory on the day France needs it and optional on the day Latvia does.
The guarantee is a common asset. Defending Latvia carries a cost. But visibly abandoning Latvia marks down the value of every promise made under the alliance.
France does not need to value Latvia as highly as Latvia values Latvia. It needs to care about the price of French commitments.
This is also why a small member can matter far beyond its military size. The question is not merely whether Bangladesh is worth defending. The question is what happens to every other guarantee after Bangladesh is not defended.
The primary function of a defence pact is not to rescue a member after an invasion. It is to enter the room before the invasion is approved.
There is an argument on the other side.
A defence pact can make a country more secure. It can also make that country more strategically visible. Permanent Pakistani combat bases in Bangladesh, for example, would alter India’s threat perception. India could respond with more deployments, more pressure and closer military cooperation with other powers.
An arrangement presented as defensive in Dhaka may look different from New Delhi. This is the security dilemma. One country buys a shield. Its neighbour sees preparations for a sword.
But the opposite argument is often pushed too far. Bangladesh does not become geopolitically invisible by staying outside an alliance either. The Bay of Bengal does not move. The Indian border does not disappear. Chinese companies do not stop building infrastructure. Myanmar does not become stable. Bangladesh cannot opt out of geography. It can only decide how to manage it.
There is also a large distance between no defence relationship and a permanent Pakistani or Turkish base in Chattogram.
That middle contains intelligence sharing, joint exercises, common procurement, air-defence cooperation, cyber support, maritime surveillance, temporary deployments and contingency planning. Bangladesh can negotiate some of these and reject others.
The proper question is not whether joining raises tension at all. Almost any meaningful security policy changes somebody’s calculation. The question is whether the deterrence gained from the arrangement reduces the probability of coercion and war by more than the alignment itself increases it.
A badly designed pact could make Bangladesh an outpost in someone else’s competition. And a well-designed one would make Bangladesh harder to isolate without turning it into anybody’s forward base.
Neutrality may also be a rational strategy. But neutrality is not a cloaking device. It requires military capacity, diplomatic balance, economic resilience and enough internal cohesion that foreign powers cannot purchase competing parts of the political system.
Then comes the economic objection.
Every taka spent on defence has a more immediately appealing use. It could buy LNG cargoes, subsidised rice, libraries, public hospitals, electricity support, cold-storage facilities, school meals or another ambitious programme to train freelancers.
This arithmetic is not false. It is incomplete.
A household could cancel its fire insurance and spend the premium on a better refrigerator. The refrigerator will improve life immediately. The insurance may produce nothing visible for twenty years.
People still buy the insurance because the event it covers, though unlikely, is large enough to ruin them.
Military spending works on the same category of risk, with one important difference. Insurance mostly compensates after the loss. Defence is supposed to reduce the probability of the loss in the first place. It is the insurance, alarm, sprinkler and fire brigade all bundled together.
Successful deterrence looks unproductive. Its output is an invasion that did not happen, a blockade that was not attempted, a factory that stayed open and a port that continued unloading cargoes.
No ministry gets to cut a ribbon at the inauguration of an avoided war. Such a goalkeeperish job.
That does not mean all military spending is sensible. Governments can overpay for the wrong equipment, buy prestige systems with little relevance, or sign a defence agreement full of exclusions with partners that cannot physically deliver help.
A bad insurance policy remains a bad policy.
But the choice is not between military spending and costless peace. The choice is between different ways of managing a low-probability event capable of damaging the whole economy at once.
Bangladesh’s own military remains the deductible. It has to survive the opening stage, protect critical infrastructure and make an attack costly from the first day.
The alliance is closer to catastrophe reinsurance. It gives Bangladesh possible access to things that are expensive to reproduce alone: strategic intelligence, early warning, advanced air defence, electronic warfare, strategic transport, naval surveillance, emergency ammunition and diplomatic coordination.
Bangladesh would not inherit the Turkish, Pakistani and Saudi militaries overnight. It would gain an option on selected capabilities, and the prospective attacker would have to price the chance that those options are exercised.
That is why alliances can be economical for smaller states. Intelligence networks, surveillance aircraft, missile warning, military transport and weapons research have high fixed costs. Sharing even part of that machinery can produce more deterrence per taka than buying every layer of strategic autonomy at retail price.
It is not free security either. Bangladesh would still have to spend on compatible equipment, exercises, communications, logistics and readiness.
The economic claim is not that the military budget automatically falls. It is that each taka may buy more deterrence.
The civilian alternatives are not irrelevant to national security either. LNG supply, strategic fuel stocks, rice reserves, cold storage, grid resilience, ports and hospitals all determine whether the country can survive a crisis.
A cold store cannot shoot down a missile. An air-defence battery cannot keep potatoes fresh. A serious security policy may need both.
The point is not to maximise military spending. It is to pay the lowest credible premium against an event capable of destroying the ability to fund everything else.
Deterrence can also affect ordinary negotiations.
It does not make Bangladesh equal to India. Geography, market size and military power do not disappear because a treaty is signed. What it can change is the disagreement point: what each side believes will happen if negotiations fail.
Suppose Bangladesh and India are negotiating transit, water-sharing, border infrastructure or access to a port. If India believes military coercion remains cheap and bilateral, that possibility sits quietly in the background even when nobody mentions it.
A credible defence relationship makes that option more expensive.
Bangladesh still does not negotiate as an equal. But its refusal may carry more weight, because the larger neighbour has fewer cheap ways to turn that disagreement into compliance.
The same logic applies to energy contracts, rail links, digital infrastructure and market access. Security does not replace economic diplomacy. It changes the floor on which the negotiation takes place.
The pact could also reduce trust and lead India to use more non-military pressure. There is no free strategic lunch here. The value depends on whether Bangladesh can raise the cost of coercion without needlessly turning every commercial dispute into a military one.
Bangladesh should not sign a defence pact just because it contains an Article 5-style promise.
The real questions come afterwards. Who shares intelligence? Who can move aircraft, troops and supplies, and how quickly? Which ports and airfields can be used? Are the communications systems compatible? What happens when a runway is damaged or a shipment is delayed?
Those details matter more than the wording of the treaty. A pact without exercises, stockpiles, access arrangements and a workable reinforcement plan may look impressive on paper and still offer little when a crisis begins.
Bangladesh should also negotiate for maintenance facilities, local production, technology transfer and reliable wartime resupply. “Defence cooperation” should not mean little more than buying equipment from three different catalogues.
The pact may never be tested. Allied aircraft may never arrive, and no emergency convoy may ever leave port. That would not necessarily mean it had failed. Its value may lie in making the military option so costly and uncertain that it is never attempted.
The pact does not have to win Bangladesh’s war. It has to help ensure that war never starts.
—
Taukir Aziz is a finance professional and a trustee of Panam Institute, a Dhaka-based think tank