Is a Third World War to be feared and what could it look like?

Since 1945 and the first use of nuclear explosives, a conviction has shaped Western strategic thinking: the existence of these “absolute weapons” makes any war of conquest between major powers unthinkable, rendering the territory of states armed with nuclear weapons inviolable. These countries could only confront each other indirectly, in limited wars, whose intensity would never reach the hyperbolic violence of the first two world conflicts.

However, this certainty has been shattered. By invading Ukraine, a country whose independence and security it had guaranteed in the Budapest Memorandum in 1994, Russia used its atomic arsenal as a shield (without risking direct involvement of the United States) to wage a conventional war of conquest. This Russian invasion has caused a profound disruption of deterrence mechanisms, the consequences of which may not have been fully diagnosed.

The concept of a “nuclear threshold,” theorized in 1960, assumed a precise line beyond which nuclear war became certain. Since the beginning of the war in Ukraine, this notion can no longer be strictly understood. In reality, behaviors obey more complex mechanisms: there is a zone of uncertainty, an intermediate space where an infinite number of hostile acts remain possible without automatically leading to the ultimate escalation.

In other words, there is an increase in the threshold beyond which the behavior of certain actors becomes intolerable. This increase opens an opportunity for “revisionist” powers, wishing to modify the rules of the system in their favor.

For example, by using force to annex new provinces and disregarding a cardinal principle of the United Nations: the inviolability of borders. According to this principle, borders cannot be changed by force, and any modification must be made within existing administrative limits. This principle had only seen rare exceptions in seventy years (Tibet acquired by China in 1950, Kashmir, the border between the two Koreas, Arab-Israeli wars, Northern Cyprus).

The return of wars of conquest

We are witnessing the most serious risk: not a “Third World War” deliberately declared by a power or group of powers, leading to total atomic war, but a multiplication of simultaneous conventional conflicts exhausting American capabilities and will to respond, which could be called a sub-threshold world war (not initially provoking the use of nuclear weapons).

For the past five years, the most significant ruptures have been caused by nuclear powers themselves. Russia attempted to subdue Ukraine through a lightning offensive and formally annex five provinces, before settling into a protracted conflict with lasting consequences for European order. Israel, an undeclared nuclear power, responded to the Hamas attack on October 7, 2023, with unprecedented military operations in Gaza, Lebanon, Syria, against Houthi rebels in Yemen, and Iran, according to its doctrine of “disproportionate response.”

The United States, far from being spectators of the system’s deregulation, have become one of the agents. The operation in Iran was launched without a UN mandate or consultation with the US Congress. Washington openly threatens NATO members, thus undermining the institutions that the US had helped establish. The guarantor of the previous order, tired of financing the Alliance, has launched a brutal reform that disrupts its architecture and threatens to destabilize it.

Other conflicts, without involving nuclear arsenals, have been initiated in correlation with these confrontations. In September 2020, Azerbaijan launched its first victorious offensive on Armenia, leading to the gradual disappearance of the Republic of Artsakh (Nagorno-Karabakh) and the exile of over 100,000 Armenians, without the international community being able to prevent this exodus.

This conflict is added to the wars between Cambodia and Thailand, India and Pakistan, or Pakistan and Afghanistan. These situations show that the return of limited local wars was not accidental, but a significant trend, in addition to the insurgency battles of previous decades.

Indeed, these wars have not all led to significant border changes, but neither the US nor its strategic competitors are able to regulate all of them at once. The US could once balance all regions and tensions through external intervention (traditionally called “offshore balancing”), but the multitude of urgent situations and conflicts no longer allow them to act sufficiently with the same budget. The multiplication of conflicts shows that this has become much more difficult. This leaves much more room for local actors to alter their relationships with neighbors.

In particular, Russia has shown other revisionist powers that economic sanctions can be absorbed, that Western war efforts have industrial and political limits, and that nuclear protection offers a broader range of conventional action than previously thought. All of this constitutes an “incentive” in the precise sense that game theory gives to the term: an increase in reward for an action, or a reduction of risk, to use force to reshape territory and power balances. To the point of questioning the very nature of the international system?

When conflicts threaten to merge

Raymond Aron, in the book “Les Guerres en chaîne,” published in 1951, noted that American strategists immediately after the war had only considered two scenarios: armed peace without direct confrontation or total war leading to nuclear detonation. He believed they omitted a third scenario, that of “limited hot wars,” such as the Korean War (1950-1953), which surprised America.

However, despite the terrible losses they sometimes caused, none of these “hot wars” degenerated into a conflict involving two coalitions directly intervening. External interventions, such as those of the Soviet Union and China in support of North Vietnam, had to be discreet, or limited to defensive aid, aimed at protecting the ally’s territorial integrity.

Our era is perhaps reinstalling conflicts “in series”: they are now interconnected, so that each new hotspot amplifies previous ones and increases the burden on the entire system.

Nuclear deterrence, until now, had confined local wars to the territory of the states involved. But the multiple, distributed, intense conflicts we are witnessing, which we cannot stop, have taken on such a scale that a possibility has emerged: the creation of an integrated conflict chain (or more precisely a concatenation), where all “local hot wars” produce a single and uncontrollable conflict, modeled after the world wars of the 20th century.

To use an analogy from the field of electricity, during the Cold War (1947-1991) and the period of US power monopoly, conflicts “derivated” off the international circuit. Each could flare up or die out independently of the others, without disrupting the overall system. A short circuit in one place did not affect the rest.

Our era is perhaps reinstalling them “in series”: conflicts are now interconnected, so that each new hotspot amplifies previous ones and increases the burden on the entire system.

What would happen if a significant number of conflicts were to occur in series? No power would be able to regulate local conflicts through a sufficient projection of power.

The US military, theoretically designed to conduct two major wars simultaneously, can practically only engage in one at full intensity. The US pays the consequences of its global dominance: it must be strong in all areas, while each adversary only needs to dominate its own region.

This structural asymmetry is at the heart of the risk of making conflicts serial: an additional crisis in Taiwan, the Gulf, or the Indian subcontinent would suffice to put Washington in a situation of strategic overload, unable to contain all focal points of tension simultaneously.

Europe, still undecided on the path to take, between loyalty to the transatlantic link and strategic autonomy, is not united enough to replace the United States. China, despite its undeniable rise in power, lacks the means (high-seas fleet, sufficient foreign bases) and the will to intervene in conflicts (its discretion in the Iran conflict shows that it prefers others to solve issues related to its hydrocarbon supply at lower costs).

A policy of passive security, based solely on the existence of nuclear arsenals and defensive alliances, is no longer enough to protect democracies. Therefore, they must contribute, through their affirmation, to build new regulatory mechanisms capable of keeping conflicts in derivation, i.e., preventing them from merging. This not only requires restoring shared norms on the use of force but also constructing a new security regime based on regional power balances that can function without depending on a single, increasingly erratic guarantor.

The end of the “indispensable nation” is written into the rebalancing of the gross domestic products (GDPs) of major powers. The international system is undoubtedly shifting toward another model. But this could take two very different forms: either it will be more distributed and polycentric, what political scientist Jean Baechler called the “oligopoly world,” or it will be a new bipolar system organized around Washington and Beijing. It remains to be seen whether, in the interim period, the world will be shielded from a breakdown of deterrence and conflict management, which would lead back to a period of uncontrollable violence similar to 1914-1945.

A policy of passive security, based solely on the existence of nuclear arsenals and defensive alliances, is no longer enough to protect democracies. It is therefore necessary for them to contribute, through their affirmation, to build new regulatory mechanisms capable of maintaining conflicts in derivation, i.e., preventing them from merging. This not only requires re-establishing shared norms on the use of force but also constructing a new security regime, based on regional power balances capable of functioning without relying on a single, increasingly erratic guarantor.

Antony Dabila is a research associate at Sciences Po’s Center for Political Research (Cevipof) and holds a doctorate in historical sociology of politics, specializing in international relations, strategic studies, and contemporary changes in democratic institutions in the face of geopolitical reshuffles.