Join War on the Rocks and gain access to content trusted by policymakers, military leaders, and strategic thinkers worldwide.
Gaza did not have to look the way it looks. That is not a moral claim. It is an operational one.
Claims of necessity are invoked to explain the scale of civilian harm, but they founder in the face of operational logic. Supporters of Israeli methods argue that Gaza’s urban battlefield — tunnels, rocket fire, extreme population density, hostages held underground — left no viable alternative to large-scale destruction. That argument rests on a false assumption: that large-scale fires were inevitable rather than chosen, and that the only tradeoff was between more fires and mission failure. That is a false choice, and I’ll explain why.
I have planned and fought in urban operations, where the tension between necessity and mission is real. Not only were better alternatives available, but the methods employed in Gaza actively reduced Israel’s chances of achieving its stated objectives: recovering hostages and dismantling Hamas. The same conditions cited to justify large-scale destruction are precisely the conditions that make discrimination, separation, and intelligence-driven operations necessary. My argument builds on a previous article published in these pages and responds directly to the critics who challenged it.
In international humanitarian law, military necessity permits measures essential to achieve a legitimate military purpose. But the standard does not ask how much force can be applied. It asks which methods are required to accomplish the mission. It is a comparative standard: whether the means chosen were necessary to achieve the objective, and whether feasible, less harmful alternatives were available.
The question is therefore whether the scale and method of force employed in Gaza were required to achieve Israel’s stated objectives. They were not. The claim that military necessity compelled the approach used does not hold, and the most direct way to demonstrate that is to examine the objective most often described as paramount: the recovery of the hostages.
The presence of hostages should tighten how force is applied. Hostages convert the battlespace into a sensitive-site problem at scale. Heavy fires in such environments produce predictable effects: collapsing structures, secondary explosions, suffocation, loss of intelligence, and forced movement of captives.
A hostage-focused approach would have emphasized intelligence-driven isolation of the battlespace, controlled cordon and search operations, sustained surveillance, and targeted raids against holding elements, reinforced at the policy level through negotiated pauses and exchanges. Fires would have been relegated to tightly constrained supporting roles. None of these methods would have guaranteed success. That is not the standard. The question is whether alternative approaches existed that were better aligned with the priority task and carried less predictable risk to the hostages.
Critics argue that the scale and dispersion of hostages made such an approach unrealistic. But hostage recovery operations are rarely conducted under ideal conditions. The challenge is to create the conditions under which intelligence can be developed and exploited, not to substitute firepower for intelligence.
Israel possessed the tools to do exactly that: advanced intelligence capabilities and a world-class hostage-rescue capacity. It also had direct access to U.S. special operations personnel providing intelligence and technical support for hostage recovery. Yet it chose to proceed with a large-scale offensive. As the campaign shifted toward intensive fires and ground invasion, hostage recovery became increasingly secondary in practice — if not in rhetoric — to the objective of destroying Hamas.
The method adopted was poorly suited to recovering living captives. It degraded intelligence, drove dispersal and concealment, increased the risk of fratricide, and substituted territorial control for the recovery of the hostages. Once hostages are killed or moved beyond reach, no amount of cleared terrain reverses that outcome.
If the methods employed were poorly suited to recovering hostages — the objective most often described as paramount — the necessity argument collapses at its most important test.
The same problem runs through Israel’s second objective: dismantling Hamas as an organization. Israeli forces never took control of the civilian environment in which Hamas was operating, and that shaped everything that followed.
Civilian movement, aid distribution, and population concentration remained largely unmanaged. Units operated in close proximity to civilians moving unpredictably through the battlespace — making identification harder, slowing clearance, and pushing greater reliance on fires to manage risk. As one assessment notes, large-scale humanitarian demands in urban combat can “rapidly unhinge the tempo and capacity of a formation.”
Separating civilians from the battlespace is a basic requirement for controlling a fight against an embedded adversary. Without it, Hamas retained concealment, access to resources, and freedom of movement, while shaping the narrative around the use of force. The population did move during the campaign, often in large numbers. The problem was that movement was never controlled in a way that supported operations.
There was space to do this. Coastal terrain along the Mediterranean offered open ground where protected areas could have been established. Civilians ultimately concentrated there during the conflict. What was missing was operational design.
The scale — 2.3 million people — is real but misframes the problem. The answer was never to concentrate the entire population in a single location. A workable approach would have established secured zones distributed along the coastal strip, each absorbing a manageable segment of the population. Consecutive adjacent zones, established as ground operations progressed sector by sector, would mean the Israeli military never had to manage the entire population at once. If one area came under pressure, the others would continue to function.
Movement into those zones would have been phased and tied to ground operations — civilians directed along designated routes during set time windows, controlled by military police and reserve units managing traffic points, regulating flow, and preventing movement back into contested areas. Controlled routes allow screening, generate intelligence, and limit adversary movement.
Entry would have been screened. Not perfectly — but screening improves over time and degrades Hamas’s freedom of movement. Israel already had the tools: biometric registration, identity checks, and data systems capable of tracking movement and flagging irregular patterns. The Israel Defense Forces operate such systems in the West Bank as a matter of routine.
These tasks — route control, checkpoint operations, screening, and sustainment — are not the work of frontline maneuver units. They are handled by military police, engineers, territorial units, and reserves. Israel mobilized large reserve formations at the outset of the campaign and has long maintained the institutional capacity to manage population movement at scale. The constraint was not a shortage of forces. These functions were never treated as a central operational requirement.
Control of aid was inseparable from operations. Hamas has long used distribution networks to maintain authority and sustain its presence inside civilian areas. In Gaza, distribution took place under uncontrolled conditions — creating disorder, exposing civilians to risk, and allowing Hamas to retain influence over the population. Treating humanitarian activity as separate from operations left those conditions unmanaged.
These zones would have operated under threat from indirect fire, infiltration, and disruption. But defined perimeters, controlled entry points, and overwatch from standoff positions constrain freedom of movement and make it harder for armed groups to operate freely inside them. Without controlled areas, Hamas operates across the entire population. With them, its movement is constrained, its access to resources reduced, and its presence easier to identify.
What developed instead left civilian movement uncontrolled, aid distribution contested, and large populations intermingled with Hamas fighters across the battlespace — an environment that worked entirely in Hamas’s favor.
The campaign relied on extensive fires across densely populated terrain, followed by ground clearance — the bluntest tool available, and among the least effective. Broad destruction degraded intelligence, impeded movement and target identification, and had limited effect on underground networks. It left the population intermingled and the operating picture degraded.
A campaign aligned with dismantling Hamas would have emphasized separation through secured corridors and protected zones, sustained intelligence collection, and targeted operations against leadership, tunnel nodes, and command elements. These methods disaggregate an embedded organization from the population it depends on. Massed fires do not.
Some critics argue that the problem runs deeper than human shielding in the conventional sense. Hamas, they contend, did not merely use civilians as cover: It designed its entire defensive posture around the expectation that civilian casualties would follow regardless of what Israel chose to do. Restrain, and Hamas operates freely among the population, preserving its fighters, its tunnels, and its command structure. Strike, and Hamas wins the legitimacy battle, converting Israeli firepower into its most effective recruiting and political tool. Geoffrey Corn has described this as a more aggravated form of civilian exploitation: not shielding in the traditional sense, but baiting — engineering conditions in which the attacker becomes an instrument of the defender’s strategic objectives. This is the strongest version of the critics’ case, and it deserves a direct answer.
The dilemma only holds if the civilian population remains intermingled with Hamas fighters. Population separation undermines that condition directly. Screen civilians into controlled zones, deny Hamas access to the human terrain it depends on, and the bait and bleed strategy loses its leverage.
Measured against Corn’s own framework, Israel’s actual response represents the worst available option. An attacker who responds to deliberate civilian exploitation with large-scale fires hands the defending force exactly the strategic victory it was seeking. Mass fires simply capitulated to Corn’s dilemma.
The harder objection is that separation takes time, during which Hamas retains the ability to exploit the civilian population. But this is precisely where the intelligence picture matters. Before the campaign began, Israeli and U.S collection assets were conducting operations that were, by some accounts, beginning to produce results: mapping networks, tracking movement, and developing the picture of where hostages were held and how Hamas’s command structure was organized. The decision to launch a large-scale offensive interrupted that process and forced Hamas to disperse. Fighters scattered, hostages were displaced, and a network that had been partially mapped dissolved and reconstituted elsewhere. The offensive produced the exact outcomes it was meant to prevent.
The period required to establish those zones would not have been a liability. As civilians moved into controlled areas, the human terrain would have simplified, clearer patterns would have emerged, and the targeting picture would have sharpened. Separation creates the conditions intelligence requires.
None of this would have been easy. Separating a population from an embedded adversary under combat conditions is one of the hardest problems in urban warfare. But the choice was between an imperfect approach that erodes Hamas’s leverage over the civilian population and one that delivered the strategic victory Hamas had designed from the outset.
Another defense of the campaign is that Hamas’s tunnel network fundamentally altered the fight. Examined closely, that logic does not hold.
Bombarding buildings to collapse tunnels below is among the least effective ways to counter an underground network. Tunnels are designed to withstand surface destruction. Heavy bombardment destroys infrastructure while obscuring the intelligence needed to locate and exploit tunnel entrances. Israeli after-action analysis confirms this: As terrain was reduced to rubble, units struggled to identify targets, coordinate fires, and maintain boundaries between maneuver elements. The resulting environment — what Israeli analysts termed “devastated terrain warfare” — degraded situational awareness and increased the risk of fratricide. The method did not merely fail to solve the operational problem. It made it harder to accomplish.
Effective counter-tunnel warfare depends on intelligence, mapping, pattern-of-life analysis, and controlled access — not massed fires. Once hostages are factored in, the argument weakens further. If captives are held inside tunnels, collapsing them is not a rescue method. There is no plausible way to recover living hostages without entering the tunnel system — and the Israel Defense Forces had the tools to do so. It has invested for years in counter-tunnel capabilities: drones, sensors, robotics, and specialized engineering forces trained to locate, breach, and clear subterranean networks. Tunnel warfare is slow and manpower-intensive. That is the nature of the problem.
Paradoxically, the tunnel network strengthened the case for population separation. If Hamas fighters were primarily underground, the decisive problem above ground was denying them freedom of movement, logistics, and replenishment. Massive surface destruction made this harder — erasing reference points, degrading intelligence, and creating conditions in which an underground adversary could operate more effectively.
Gaza’s operational complexity did not require the method chosen. It made choosing the right method more important.
Sustained rocket fire against Israeli communities was a real threat, but it did not require large-scale devastation of Gaza’s urban terrain. The most effective means of suppression are detection, interception, and disruption of launch activity — capabilities Israel possessed at scale.
Israel fields one of the most sophisticated layered air defense architectures in the world. Systems such as Iron Dome — complemented by radar coverage, airborne surveillance, and multi-source cueing — have proven effective at intercepting rockets and mitigating their impact.
Rocket launches are not random. They depend on observable behaviors: preparation, movement of teams, emplacement of launch systems, communications, and the repeated use of certain areas. Persistent surveillance — drones, aerostats, manned aircraft, and signals intelligence — combined with rapid precision strikes against launch teams and equipment, are historically the most effective suppression methods.
Large-scale destruction undermines these efforts. Rubble creates concealment, obscures movement, and degrades pattern-of-life analysis, providing small, mobile launch teams with more opportunities to hide, maneuver, and evade detection.
The more durable solution is to control surface terrain, isolate launch zones, and deny freedom of movement — again pointing toward separation, secured areas, controlled corridors, and intelligence-led targeting.
The effectiveness of warnings has been addressed elsewhere, but remains central to defenses of the campaign and warrants brief reconsideration. The relevant question is not whether warnings were issued, but whether they reduced harm. On that standard, the record is mixed.
Communications blackouts often prevented civilians from receiving or acting on warnings, while evacuation orders were frequently vague or directed movement without secured routes or viable destinations. In some cases, areas identified as safer locations were later struck. Warnings are only protective when civilians can realistically comply and expect reduced risk. Without secured corridors and genuinely protected areas, warnings shift risk onto civilians while preserving a claim of procedural compliance. That is risk transfer, not mitigation.
Some defenders of the campaign point to the ratio of civilian deaths per munition as evidence of restraint. In a campaign involving tens of thousands of strikes, that metric is uninformative. Even a low average casualty rate per strike produces very high aggregate harm. The statistic says nothing about whether targets were appropriately selected, whether intelligence was sufficient, or whether the anticipated military advantage justified the harm.
This problem is clearest in how the battle of Mosul is invoked. Casualty figures from that campaign are frequently cited to suggest that outcomes in Gaza fall within the historical range of modern urban warfare. That comparison misuses the underlying data and ignores critical differences in how the campaigns were fought.
The figures commonly cited for Mosul — often ranging from several thousand to more than 9,000 deaths — are aggregate totals drawn from a complex battlespace involving coalition airstrikes, Iraqi ground operations, Islamic State violence, and a significant number of indeterminate cases. Attribution is frequently unclear or contested. A substantial but unquantified share of civilian harm — including deaths caused directly by Islamic State forces — remains embedded within these totals. They cannot be disaggregated into meaningful ratios or attributed to any single actor.
The tactical comparison is equally flawed. Mosul was a high-intensity urban battle against a terrain-holding force that fought above ground from fortified positions and could not be bypassed — compelling Iraqi units to conduct deliberate block-by-block clearing under constant fire from snipers, improvised explosive devices, and ambushes. Hamas fighters in Gaza operated differently: dispersed, relying on concealment, ambush, and withdrawal, rarely mounting sustained above-ground defenses. An adversary that avoids decisive engagement above ground does not impose the same requirement for large-scale fires.
The capability gap matters equally. Iraqi forces lacked independent surveillance and reconnaissance assets, intelligence fusion, and precision strike capacity, relying heavily on coalition support. Military necessity is capability-dependent: What is necessary for a force operating with limited situational awareness is not necessary for one equipped to reduce uncertainty. Israel operated with persistent surveillance, precision-guided munitions, unmanned systems, and highly networked intelligence — capabilities that expand the available options for distinguishing combatants from civilians. Yet the Gaza campaign relied heavily on standoff fires across densely populated terrain, often before sustained ground contact.
Treating casualties produced under these fundamentally different conditions as interchangeable obscures the relevant question. The issue is not how outcomes compare in the aggregate, but whether the methods used were required and whether less harmful alternatives were available.
The central question was never whether Gaza was a challenging environment. Nor was it whether Hamas posed a serious threat, employed tunnels, took hostages, fired rockets, or embedded itself among civilians. All of that is true. The question was whether those conditions compelled the methods chosen — or whether they were invoked to justify choices that were poorly aligned with Israel’s own stated objectives.
Examined operationally, the answer is clear. The recovery of hostages, the dismantling of Hamas as an organization, and the protection of civilians all demanded discrimination, separation, intelligence dominance, and controlled application of force. Instead, the campaign relied on extensive fires, permissive targeting under uncertainty, and large-scale destruction in dense terrain. Those choices predictably transferred risk onto the civilian population, degraded intelligence, drove Hamas to disperse and adapt, and undermined the very objectives military necessity was invoked to defend.
Military necessity is a comparative standard: whether the measures employed were required to achieve a legitimate military purpose, and whether feasible, less harmful alternatives existed. Such alternatives were available. They were not risk-free, but they were better suited to recovering hostages, disaggregating Hamas from the population, and preserving the intelligence required to dismantle an embedded organization.
Gaza’s exceptional difficulty did not require the method chosen. It made choosing the right method more important.
Andy Milburn is a retired U.S. Marine Corps officer. He fought in Fallujah, led an infantry battalion in Iraq’s Anbar Province, and commanded a special operations task force during the campaign to retake Mosul from the Islamic State. He later led a humanitarian organization operation on the front lines in Ukraine. He is the author of When the Tempest Gathers: A Marine Special Operations Commander at War.
Image: Jaber Jehad Badwan via Wikimedia Commons
