The Legal Limits of Urban Infrastructure Bombardment
On the night of Feb. 13, 1945, Allied bombers struck Dresden in successive waves. The stated logic was military-industrial: Dresden was a major rail junction feeding the Eastern Front, a communications hub, and a center of war-related manufacturing. The result was the destruction of the city center, causing approximately 25,000 deaths.
Dresden has since become shorthand for a particular mode of warfare, one that treats an entire city as an integrated military-industrial system—where rail yards, factories, electrical networks, and communications nodes are struck not as discrete objectives but as interconnected elements of an enemy’s capacity to sustain operations. As Richard Overy documented in The Bombing War: Europe 1939–1945, Allied strategic bombing doctrine evolved from targeting specific factories toward area bombardment of urban centers on the premise that destroying workers’ housing, transport links, and utilities would cripple war production. Overy’s assessment is sober: Precision was elusive, the strategy was often ineffective, and civilian death tolls were staggering.
This article does not adjudicate whether Dresden was lawful under the standards prevailing in 1945—the law of aerial bombardment was underdeveloped, and that assessment belongs to a different era. What Dresden illustrates is a recurring strategic impulse: treating a city’s infrastructure as a single target set rather than as discrete objects requiring individual legal justification.
Modern international humanitarian law imposes strict conditions on such an approach, replacing system-level targeting logic with object-level legal obligations. Even where dual-use infrastructure is present—and it almost always is—the framework demands granular discrimination, individually articulated military advantage, and feasible precautions for each attack. This article clarifies those legal conditions, stress-tests them against the NATO air campaign over Kosovo in 1999, and considers implications for contemporary urban warfare.
The Modern Legal Framework Governing Bombardment
The conduct of hostilities rests on four interdependent principles: distinction, proportionality, military necessity, and precautions in attack. Military necessity permits only the degree and kind of force required to achieve the legitimate purpose of a conflict—it authorizes action but simultaneously constrains it. These principles derive primarily from Additional Protocol I (AP I) to the 1949 Geneva Conventions and from customary international humanitarian law as identified in the International Committee of the Red Cross (ICRC) 2005 study. Together, they establish boundaries that apply in both international armed conflicts and, through customary international humanitarian law, in non-international armed conflicts.
The principle of distinction, codified in Article 48 of AP I, requires parties to an armed conflict to distinguish at all times between the civilian population and combatants, and between civilian objects and military objectives. It is the foundational rule from which the other targeting obligations flow.
Article 52(2) provides the operative definition of a military objective: an object which by its nature, location, purpose, or use makes an effective contribution to military action, and whose total or partial destruction, capture, or neutralization offers a definite military advantage in the circumstances ruling at the time. Both conditions—effective contribution and definite advantage—must be satisfied simultaneously. An object that meets only one prong of the test retains its civilian character.
This definition is reflected in the U.S. Department of Defense Law of War Manual (§ 5.7.3), which requires targeting decisions to be made in good faith and establishes a presumption that persons or objects are protected unless available information indicates otherwise (§ 5.4.3.2).
The dual-use problem is immediately apparent. A power grid feeds both military command centers and civilian hospitals; a bridge carries military logistics and civilian traffic. Under Article 52(2), the question is not whether the object has any military function, but whether its specific military contribution and the concrete advantage of its destruction satisfy both prongs of the test. As the ICRC has emphasized, the mere possibility of military use is insufficient; the contribution must be assessed on the basis of reliable information indicating that the object effectively serves military purposes at the relevant time.
Article 51(4) of AP I prohibits indiscriminate attacks—those not directed at a specific military objective, those employing a method or means of combat that cannot be directed at a specific military objective, or those whose effects cannot be limited as international humanitarian law requires. Article 51(5)(a) specifically prohibits bombardment that treats as a single military objective a number of clearly separated and distinct military objectives located in an area containing a concentration of civilians or civilian objects. This is the prohibition on area bombardment—a deliberate rejection of treating concentrations of military objectives within cities as a single targetable entity.
The prohibition does not mean that multiple military objectives in an urban area cannot be struck. It means that each must be individually identified and targeted. Treating an infrastructure network as a single target set, on the rationale that its components are functionally interconnected, risks precisely the kind of indiscriminate attack that Article 51(5)(a) was designed to prevent.
The rule of proportionality, codified in Article 51(5)(b) of AP I and recognized as customary international humanitarian law (ICRC Customary IHL Rule 14), prohibits attacks expected to cause incidental civilian harm that would be excessive in relation to the concrete and direct military advantage anticipated. The assessment is prospective, based on information reasonably available to the decision-maker at the time of the attack—not with the benefit of hindsight. The military advantage must be concrete and direct—speculative or general advantages do not suffice. It must also be assessed in relation to the specific attack under consideration, not the broader aims of the campaign as a whole. And in dense urban environments, foreseeable incidental effects—including those resulting from the disruption of essential services—may extend far beyond the immediate blast radius.
Article 57 of AP I and ICRC Customary IHL Rule 15 require those who plan or decide upon an attack to take all feasible precautions to verify that objectives are military objectives, to choose means and methods that minimize incidental civilian harm, and to refrain from launching an attack expected to be disproportionate. Operationally, this means a joint targeting working group—as described in JP 3-60, Joint Targeting—works through each stage of the targeting cycle in accordance with established joint doctrine: target development, weaponeering, collateral damage estimation, and legal review before recommending weapons release. Where intelligence, surveillance, and reconnaissance coverage is degraded or incomplete, the obligation to verify does not disappear; it constrains the attack until adequate verification is feasible.
Rules of Engagement and Coalition Targeting Practice
In coalition operations, rules of engagement introduce an additional operational layer between legal authority and weapons release. Rules of engagement are directives issued by competent authority that define the circumstances, conditions, and limitations under which forces may initiate or continue the use of force. They translate the political will of participating states, the legal framework of international humanitarian law, and operational requirements into actionable constraints for military commanders.
In NATO operations, rules of engagement are approved by the North Atlantic Council and technically prepared by the Military Committee and NATO Strategic Commands. However, each member state retains the right to apply national caveats—additional restrictions reflecting domestic legislation, constitutional requirements, or political constraints. In practice, this means that contributing nations operating under the same campaign plan may apply different thresholds to the authorization of force against identical target sets. A dual-use infrastructure target that one nation’s targeting cell approves may be restricted or excluded by another’s national caveat.
This inconsistency represents one of the enduring challenges of multinational coalition warfare. It is directly relevant to infrastructure campaigns, where the scope of permissible targets and the degree of acceptable collateral damage may vary across contributing forces. The Kosovo campaign, examined below, illustrated this friction in practice: National caveats influenced which targets individual allies were willing to strike, creating operational disparities within a nominally unified air campaign. NATO has long recognized this problem and has advocated for the harmonization of command, control, and coordination processes, but the tension between alliance-level targeting authority and national-level restrictions remains unresolved.
A critical doctrinal point: The United States and several other major military powers are not party to AP I. However, the core targeting principles—distinction, proportionality, and precautions—are widely recognized as customary international humanitarian law binding in all armed conflicts, including non-international armed conflicts. Common Article 3 of the 1949 Geneva Conventions provides baseline humanitarian protection in non-international armed conflicts, but as Jelena Pejic argued in the International Review of the Red Cross (2011), its scope extends beyond its text through broader customary international humanitarian law. Targeting rules in non-international armed conflicts are governed by the full body of customary rules on the conduct of hostilities.
Kosovo (1999) as a Modern Stress Test
NATO’s air campaign against the Federal Republic of Yugoslavia (Operation Allied Force, March 24-June 9, 1999) provides the most extensively documented modern test case for the legal framework governing infrastructure bombardment. Over 78 days, NATO aircraft flew thousands of sorties targeting not only fielded forces in Kosovo but also an expanding set of infrastructure targets across Serbia: electrical power stations, bridges, oil refineries, communications networks, government buildings, and media facilities.
The target list expanded considerably as the campaign progressed, growing from air defense systems and military installations to encompass dual-use infrastructure, on the rationale that degrading Serbia’s economic and communications capacity would erode the regime’s ability to sustain operations in Kosovo. Power grids were struck repeatedly using both conventional munitions and specialized carbon-fiber (“soft”) bombs. By early May 1999, these strikes had knocked out power to roughly 70 percent of Serbia, affecting water systems, hospitals, and civilian life nationwide. NATO also struck over 50 bridges, both oil refineries, and the Serbian radio and television service (RTS) building in Belgrade, killing 16 civilians.
NATO articulated several military advantages: Power grids fed military command-and-control systems; bridges facilitated resupply of forces in Kosovo; media facilities were characterized as elements of the regime’s command-and-communications architecture.
The most authoritative legal assessment came from the committee established by the UN International Criminal Tribunal for the former Yugoslavia (ICTY) prosecutor to review the bombing, which issued its final report in June 2000. The committee found no sufficient basis to recommend a criminal investigation. Its mandate concerned potential individual criminal responsibility, not the broader strategic or humanitarian assessment of the campaign. Prosecutor Carla Del Ponte accepted this recommendation, concluding there was no evidence of deliberate targeting of civilians.
However, the committee’s analysis was far from a blanket endorsement. Several findings are doctrinally significant.
On target selection, the committee noted that NATO had generally directed attacks at objects meeting the Article 52(2) definition. The power transformer stations struck on April 23, 1999, were identified as supplying the air-defense coordination network. But the committee acknowledged difficulty in assessing proportionality when infrastructure serves both military and civilian functions, noting that where the presumptive effect on civilian infrastructure is grave, the military advantage would need to be correspondingly significant.
On proportionality, the committee noted that assigning relative values to military advantage and civilian harm is difficult, and that the principle is more easily stated than applied. It observed that dual-use objects with significant civilian functions raise particular proportionality concerns—especially given reverberating effects of disrupting power, water, and communications in urban areas.
On precautions, the committee examined NATO’s minimum flight altitude of 15,000 feet, maintained to protect aircrew from Yugoslav air defenses. While this may have limited visual target verification, the committee found that precision-guided munitions and surveillance technology meant the obligation to distinguish was carried out effectively in the vast majority of cases. It noted, however, that incidents including the Grdelica Gorge train strike and the Djakovica convoy attack raised serious questions about the adequacy of precautions in specific cases.
On the RTS strike, the committee offered perhaps its most nuanced assessment. It found that if the attack on Serbian state television were justified solely by its propaganda function, its legality would be questionable—propaganda use alone does not generally satisfy the Article 52(2) requirement of effective contribution to military action. However, the committee determined that NATO’s primary aim was disrupting the military command-and-control network of which RTS transmitters formed a part, with the propaganda function being complementary rather than primary.
The ICTY committee’s report highlights the doctrinal tensions inherent in infrastructure targeting. Electrical grids can be lawful military objectives when they demonstrably serve military functions—but striking them in a manner that collapses power across an entire country tests the limits of distinction and proportionality. The absence of a criminal prosecution does not resolve the broader normative debate about the humanitarian cost of infrastructure warfare; it reflects the high threshold required for individual criminal responsibility.
Contemporary Urban Warfare and Recurring Legal Tensions
The targeting patterns visible in Kosovo have recurred at greater intensity in subsequent conflicts.
In the Syrian civil war, the siege of Aleppo (2012–2016) involved extensive strikes that affected hospitals, water stations, electricity infrastructure and other essential services. Multiple parties struck dual-use infrastructure at a scale raising questions about whether individual targeting decisions satisfied the Article 52(2) standard—or whether the pattern reflected treating the city as an integrated target system.
Since 2022, the war in Ukraine has involved repeated large-scale strikes on electrical power infrastructure—transformer stations, distribution nodes, and generation facilities—resulting in prolonged blackouts, disruption of winter heating, and strain on hospital capacity. The stated rationale has centered on degrading defense production and logistics. Civilian consequences, including blackouts affecting millions, have been widely documented.
This article does not make legal determinations about specific strikes in these conflicts—such assessments require targeting intelligence and command records not publicly available. What these conflicts illustrate are recurring analytical challenges.
First is the problem of reverberating effects. In interconnected urban systems, striking a single power station may cascade through water treatment plants, hospitals, telecommunications, and heating systems across an entire region. The proportionality assessment must account for these foreseeable secondary and tertiary effects—not merely immediate damage at the point of impact. The standard concerns effects reasonably foreseeable at the time of decision, not hypothetical downstream consequences.
Second is the challenge of cumulative impact. A single strike on a transformer station may be individually proportionate. But sustained campaigns against infrastructure raise complex questions about how anticipated effects are framed within each individual strike decision. International humanitarian law assesses proportionality per attack, but commanders planning successive strikes against interconnected systems must consider how prior damage shapes the foreseeable consequences of the next.
Third is the enduring tension between system-level targeting logic and object-level legal requirements. Military planners think in terms of systems: disrupting air-defense networks, collapsing logistics chains, degrading command-and-control capacity. But international humanitarian law requires that each object be independently assessed as a military objective, its advantage independently articulated, and proportionality assessed for each attack. This gap is where the greatest doctrinal friction occurs.
Legality does not depend on the scale of destruction alone. It depends on whether each attack is directed at a specific military objective, whether incidental civilian harm is excessive relative to the military advantage anticipated, and whether feasible precautions were taken. A campaign destroying large portions of an adversary’s infrastructure may be lawful if each strike satisfies these requirements. A single strike that does not may be unlawful regardless of limited damage.
Operational Implications
The legal framework governing urban infrastructure bombardment is not prohibitive; it is demanding. Dual-use infrastructure can constitute lawful military objectives under Article 52(2). Large-scale destruction is not unlawful per se, but its legality depends on compliance with distinction, proportionality, and precautions. Treating interconnected infrastructure as a single target system without individually assessing each object risks violating the prohibition on indiscriminate attacks.
For commanders and legal advisers, several practical implications follow.
First, targeting cells must resist aggregating dual-use infrastructure into system-level target sets that obscure the individual legal analysis required for each object. A power grid is not a single military objective; each node must be justified independently.
Second, proportionality assessments must account for reverberating effects. The incidental harm from disabling a transformer station extends to every hospital, water pumping station, and heating system that depends on it. These foreseeable consequences must be part of the proportionality calculus.
Third, the requirement for feasible precautions imposes a continuous obligation. As precision-strike capabilities and intelligence, surveillance, and reconnaissance assets improve, the scope of what is considered feasible evolves alongside available means and methods. What was infeasible at 15,000 feet in 1999 may not be infeasible today.
Fourth, the articulated military advantage for each strike must be concrete and direct—not speculative, not aggregated across an entire campaign, and not couched in the generalized language of “degrading the adversary’s war-sustaining capacity.” International humanitarian law requires specificity. In practice, that specificity must be articulated under conditions of imperfect intelligence, time pressure, and evolving operational objectives.
The distance from Dresden to the modern targeting cell is measured as much in legal development as in technological advancement. The strategic logic of system warfare persists. The legal logic of discrimination operates at the level of individual objects. The tension between the two defines modern urban bombardment—and what distinguishes lawful operations from indiscriminate attack.
