The conflict between Israel and Hezbollah has entered its most dangerous phase since the Gaza war began. Over the past week, thousands of pagers and walkie-talkies used by Hezbollah members exploded across Lebanon in attacks widely attributed to Israel, killing dozens and injuring thousands. Israel then carried out a strike in southern Beirut that killed senior Hezbollah commander Ibrahim Aqil and other members of the group’s elite Radwan Force. Hezbollah responded with rockets aimed deeper into northern Israel, while Israeli aircraft have intensified strikes against Hezbollah positions across southern Lebanon. For almost a year, both sides appeared to operate according to an implicit rule: exchange enough fire to demonstrate resolve, but remain below the threshold of the full-scale war neither side appears to want. That equilibrium is now deteriorating. The central problem is not simply that Israel and Hezbollah are escalating. It is that each side believes escalation can restore deterrence, even though the accumulated result of both sides trying to restore deterrence is making deterrence progressively less stable.
This is a classic security dilemma, but one occurring between a state and a heavily armed non-state organisation. Hezbollah began firing towards Israel after the 7 October Hamas attack and Israel’s subsequent war in Gaza, presenting its actions as support for Palestinians and pressure on Israel. Israel responded with strikes against Hezbollah fighters, commanders and infrastructure. Tens of thousands of civilians have been displaced from communities on both sides of the border. Israel therefore faces a domestic strategic problem: even if Hezbollah’s attacks remain below the level of a full invasion, northern Israel cannot return to normal while residents believe rockets and anti-tank missiles can reach their homes. Restoring those citizens becomes a political objective in its own right.
This changes Israel’s deterrence calculation. If the pre-existing equilibrium means Hezbollah continues enough attacks to keep northern communities evacuated, merely preventing a larger war is not necessarily success. Israel needs to alter Hezbollah’s behaviour sufficiently that civilians believe returning home is safe. That creates pressure to impose greater costs on Hezbollah, destroy more capabilities or push its forces further from the border. The status quo can therefore become intolerable even when it is less destructive than the alternatives. States escalate not only because existing deterrence has completely collapsed but because the equilibrium produced by deterrence no longer satisfies their minimum political objectives.
Hezbollah faces a mirror problem. Its credibility rests partly on being the most powerful member of Iran’s regional “axis of resistance” and on presenting itself as capable of imposing costs on Israel. If Israeli strikes repeatedly kill senior commanders, destroy communications systems and force Hezbollah away from the border without meaningful retaliation, the organisation risks appearing deterred rather than deterrent. It therefore has incentives to demonstrate that attacks produce consequences. Both actors consequently possess a reason to move one step higher after every successful enemy strike.
This creates an escalation ratchet. Israel conducts an operation intended to weaken Hezbollah and convince it that continued attacks are too costly. Hezbollah interprets failure to respond as damaging its credibility and retaliates. Israel interprets the retaliation as evidence that previous punishment was insufficient and escalates again. Every individual action can appear rational while the collective trajectory moves towards an outcome both actors originally wanted to avoid.
The pager and radio explosions intensify this dynamic because they attacked something deeper than individual Hezbollah personnel: organisational trust. Thousands of communications devices apparently detonated almost simultaneously, demonstrating that Hezbollah’s supply chains and internal security had been penetrated. The physical casualties matter, but the informational effect may be even larger. An organisation dependent on secure communications must now question whether other devices, suppliers or logistical networks have also been compromised.
This is supply-chain warfare. Modern military organisations depend on enormous networks of components manufactured, transported and distributed through multiple intermediaries. Security traditionally focuses on preventing adversaries from intercepting communications after equipment enters service. The Lebanon attacks demonstrate the possibility that the equipment itself can become the vulnerability before reaching the user.
The strategic implications extend well beyond Hezbollah. Governments and militaries around the world will examine whether communications hardware, drones, sensors or other electronics sourced through foreign supply chains can be trusted completely. The lesson resembles the geopolitical debate over semiconductors and telecommunications infrastructure: dependency creates attack surfaces. A device does not need to transmit intelligence to become a security risk. Its manufacturing and distribution chain can itself be manipulated.
This will strengthen pressure for trusted suppliers, domestic manufacturing and verification of critical hardware. The economic cost will be higher because secure supply chains often require redundancy and reduced dependence on the cheapest producer. Again, efficiency and resilience trade against one another. Globalised production minimises costs by allowing firms to source components wherever they are cheapest. Security policy increasingly asks whether the cheapest component remains cheap after the expected cost of sabotage is included.
For Hezbollah, the attack creates an additional internal problem because organisations function through assumptions of trust. Members need confidence that commanders, suppliers and communications systems are not compromised. Once infiltration becomes plausible everywhere, internal security can become excessively suspicious. Organisations respond through investigations, compartmentalisation and restrictions on information sharing. Those measures reduce vulnerability while also reducing efficiency.
This is the paranoia tax of intelligence penetration. If every employee may be compromised, fewer people receive information. Decisions become slower. Communication becomes more cumbersome. Initiative declines because individuals fear violating security procedures. A successful intelligence operation can therefore reduce an adversary’s organisational productivity long after the immediate attack ends.
Israel appears to possess substantial intelligence penetration of Hezbollah, demonstrated not only by the communications attacks but by its ability to target senior commanders. Ibrahim Aqil had been one of Hezbollah’s most important military figures and a senior commander in the Radwan Force. His death follows the killing of other commanders during the preceding months. Leadership decapitation can reduce organisational effectiveness, but its strategic value depends on how replaceable leaders are.
Highly institutionalised organisations survive individual deaths because command structures produce successors. Hezbollah has existed for more than forty years, possesses extensive military experience and has survived the deaths of senior figures previously. Killing commanders therefore imposes costs without necessarily destroying the organisation. The relevant question is whether Israel can eliminate leaders faster than Hezbollah can replace them while simultaneously disrupting the networks through which replacements acquire information and authority.
This is an organisational attrition problem rather than simply a casualty count. Removing one experienced commander destroys years of accumulated tacit knowledge that cannot be replaced instantly even when another person assumes the title. Repeated leadership losses can therefore reduce operational quality gradually. But if the organisation possesses deep cadres and strong institutional memory, decapitation may produce diminishing returns.
Hezbollah’s relationship with Iran makes replacement easier in some respects because it belongs to a broader regional network providing weapons, financing, training and strategic support. Iran has spent decades developing armed partners across Lebanon, Iraq, Syria and Yemen partly because these organisations create strategic depth without requiring Iran to confront Israel or the United States directly every time regional conflict occurs. Hezbollah is the most militarily capable component of that network.
This means a Lebanon war cannot be analysed purely as an Israel-Hezbollah bilateral conflict. Iran’s incentives matter. Tehran wants Hezbollah capable enough to deter Israel, particularly from attacking Iranian nuclear infrastructure. Hezbollah’s large missile arsenal effectively gives Iran a second-strike proxy capability: an Israeli attack on Iran could trigger enormous rocket attacks from Lebanon.
Iran therefore has strong reasons to preserve Hezbollah’s military strength rather than allowing it to be consumed entirely in the Gaza conflict. That may create Iranian pressure for restraint even while Tehran rhetorically supports Hezbollah’s confrontation with Israel. A proxy is strategically valuable partly because it remains available for future crises.
This produces a principal-agent problem. Iran and Hezbollah possess overlapping but not identical interests. Hezbollah is deeply connected to Iran, but it is also a Lebanese political and military organisation facing its own domestic constituency and battlefield conditions. Tehran may prefer preserving the organisation. Hezbollah may face reputational pressure to respond more strongly. Proxies provide sponsors with strategic flexibility precisely because they possess some autonomy; that autonomy also means sponsors cannot always calibrate escalation perfectly.
Lebanon itself bears much of the cost despite having limited ability to control the confrontation. The country has endured years of financial collapse, currency depreciation, political paralysis and the aftermath of the 2020 Beirut port explosion. A full-scale Israel-Hezbollah war would strike an economy and state already operating under severe institutional weakness. Infrastructure destroyed in 2006 was expensive to rebuild when Lebanon’s financial system was far healthier. Repeating that destruction today would be substantially harder.
This is the tragedy of fragmented sovereignty. The Lebanese state formally controls the country’s territory and represents Lebanon internationally. Yet it does not possess a monopoly over the military decisions capable of bringing the country into war. Hezbollah maintains an armed force more powerful than the Lebanese military in several respects and makes strategic decisions partly outside ordinary state command. Lebanese civilians can therefore bear the consequences of military choices their elected government does not fully control.
This violates one of the fundamental relationships underlying democratic accountability: the actor capable of imposing national costs should be politically accountable to the population bearing those costs. If the Lebanese government chooses war, citizens can theoretically punish the governing coalition electorally. Hezbollah’s hybrid position as political party, social organisation and autonomous military force makes responsibility much harder to allocate.
Israel confronts a different accountability problem because its government possesses formal authority but must balance several simultaneous fronts. The war in Gaza continues, hostages remain held, the West Bank remains unstable, Houthis continue threatening regional shipping and Iran demonstrated in April that direct state-to-state escalation is possible. Opening a full-scale Lebanon war would stretch military resources and create enormous civilian risk.
Hezbollah’s arsenal is far larger and more sophisticated than Hamas’s. Estimates before the current conflict frequently placed its rocket and missile stockpile at well over 100,000 projectiles, including weapons capable of reaching much of Israel. Even if Israeli air defences intercept a large proportion, saturation matters. No defensive system has infinite interceptors.
This is the economics of missile defence again. An attacker does not need every missile to penetrate. It needs enough volume that the defender’s interception capacity becomes constrained. Israel’s April success against Iran demonstrated extraordinary defensive capability, but that attack lasted hours. A Hezbollah war could involve sustained barrages over weeks.
Defence therefore buys Israel strategic time rather than invulnerability.
This makes pre-emption attractive: destroy launchers, commanders and communications before the larger barrage begins. Yet pre-emption can itself trigger the war it is intended to make less dangerous. The strategic choice becomes whether the probability of future attack is high enough that striking first improves expected security.
That is one of the most dangerous calculations governments make because the counterfactual is unknowable. If Israel attacks and Hezbollah responds massively, critics can say the Israeli strike caused the war. If Israel does not attack and Hezbollah later launches a larger assault, critics can say government ignored warning signs.
Decision-makers therefore optimise under deep uncertainty, not measurable risk.
The United States and France are attempting diplomatic solutions partly because neither side appears to possess an easy military route towards its objective. Israel wants Hezbollah forces moved away from the border, broadly consistent with the logic of UN Security Council Resolution 1701, which ended the 2006 war and called for an area south of the Litani River free of armed personnel other than Lebanese state and UN forces. Hezbollah links its attacks to the Gaza war and has indicated that fire could stop if a Gaza ceasefire occurs.
That linkage makes Gaza and Lebanon strategically connected even though they involve different organisations. A Gaza ceasefire could reduce Hezbollah’s political justification for continued attacks and create space for a border arrangement. Continued Gaza fighting keeps the northern front active.
This is another example of conflict interdependence. Solving one war does not automatically solve another, but it changes the bargaining environment. Regional conflicts increasingly operate as networks rather than isolated dyads.
The danger is that the Lebanon front may now acquire enough momentum to become independent of Gaza. Once thousands of devices explode, senior commanders are killed and rockets penetrate deeper into Israel, each side accumulates grievances requiring responses unrelated to the original reason Hezbollah entered the conflict.
Wars develop endogenous causes.
The initial cause explains why confrontation began.
Subsequent violence creates new reasons for continuing.
This is why conflicts become harder to terminate over time. Every casualty creates constituencies demanding accountability. Every destroyed home increases political pressure. Every successful operation raises expectations that further force can produce victory.
The original bargaining space narrows.
This week’s escalation therefore represents more than another exchange of fire. The informal rules governing the previous eleven months are being rewritten. Operations once considered too escalatory are becoming normalised. Intelligence attacks penetrate deeper. Senior leaders become targets. Rockets travel further.
Every threshold crossed once becomes easier to cross again.
That is escalation normalisation.
The same process occurred between Iran and Israel in April when direct state-to-state attacks became possible after decades of shadow conflict. Both sides ultimately stepped away from a larger confrontation. Lebanon may still follow the same path.
But the margin is narrowing.
The most dangerous conclusion either side could draw from previous restraint is that escalation is therefore controllable.
It is controllable until one attack produces consequences the attacker did not expect.
A rocket kills dozens of civilians.
A strike eliminates a political leader.
An intelligence operation causes much larger casualties than planned.
Iran intervenes more directly.
Israel concludes Hezbollah is preparing a mass attack.
At that point, political leaders may discover that the off-ramp they expected still to exist has disappeared.
This is why deterrence is not simply the ability to punish an opponent. Stable deterrence requires credible limits. Each side needs some understanding of which actions the other can absorb without escalating further.
The Israel-Hezbollah relationship increasingly lacks those limits.
Israel believes the previous equilibrium left northern citizens unable to return home.
Hezbollah believes accepting repeated Israeli attacks without response destroys its credibility.
Both therefore want a new equilibrium more favourable than the old one.
The problem is that there may be no mutually acceptable equilibrium available through force alone.
Israel cannot eliminate every Hezbollah rocket easily without enormous destruction across Lebanon.
Hezbollah cannot force Israel to tolerate indefinite attacks on its northern communities.
That means the eventual solution will probably require political arrangements neither side can obtain entirely through military victory.
Force can change the bargaining position.
It cannot remove the need to bargain.
The communications attacks demonstrate Israel’s extraordinary intelligence capability.
The rockets demonstrate Hezbollah retains substantial military capability.
Neither fact resolves the political problem at the border.
That is the broader lesson from this week.
Tactical superiority is not the same thing as strategic resolution.
Israel can penetrate Hezbollah’s supply chains.
Hezbollah can continue launching rockets.
Israel can kill commanders.
Hezbollah can replace some of them.
Both can impose increasingly severe costs.
The question is whether those costs change behaviour faster than they create new reasons for retaliation.
For eleven months, both sides acted as though they understood where the threshold for full war sat.
This week suggests that threshold was never fixed.
Every escalation moves it.
And eventually a strategy based on repeatedly approaching the edge encounters the basic problem of edges:
one miscalculation is enough to discover that you already crossed it.
