Balance of Power in International Relations

Balance of Power

The Confusion at the Centre of the Concept

The balance of power is the oldest continuously used idea in the study of international politics and the least agreed upon. It has been called the iron law of the states system and it has been called a slogan. The charge that the notion is notoriously full of confusion is not an exaggeration, but the confusion has a diagnosable cause, and once stated it makes the concept usable again.

• The phrase does three logically different jobs, and most quarrels about it are quarrels about which job is meant.
  • As a description it names a state of affairs — a distribution in which no single state or bloc predominates. So used, it is an empirical claim that can be true or false of a given moment.
  • As a prescription it names a policy — the deliberate prevention of any rival’s preponderance. So used, it is advice, and can be wise or foolish.
  • As a prediction it names a recurrent outcome — the tendency of balances to re-form whether or not anyone intends them, which is Kenneth Waltz’s claim.
• The three senses have different truth conditions, which is why disputes about relevance run in circles: a critic showing that no equilibrium exists today has refuted the description without touching the prediction.
• A fourth confusion sits underneath the three — the word balance is itself used in incompatible ways.
  • The equilibrium sense, from the scales: balance means A = B, an approximate equality of weights.
  • The residual sense, from accounting: a balance is what is left over, A minus B, so that “the balance lay with Britain” means Britain had the surplus — the opposite of equality.
  • The loose sense, in which the phrase means no more than “any distribution of power at all”.
• The concept is not incoherent but overloaded. Separate the three jobs and the senses of balance, and the objections sort into ones that bite and ones that miss.

The phrase used as	What it asserts	How it could be falsified
Description	No state now predominates	Show a preponderant state
Prescription	Statesmen ought to prevent preponderance	Show the policy fails on its own terms
Prediction	Balances re-form automatically under anarchy	Show sustained unbalanced power
Loose usage	“the way power is distributed”	Nothing — it is not a claim

What the Balance of Power Means

The working definition

• The balance of power is a condition in which capabilities are distributed among states or blocs such that no single actor can dominate the rest. The equilibrium is produced by states building their own strength or combining with others.
• The purpose is negative rather than positive — not peace or justice as such but the prevention of hegemony, so that peace, where it follows, is a by-product of frustrated ambition.

Morgenthau’s four meanings

Hans Morgenthau made the ambiguity explicit rather than resolving it, listing four things the phrase is used to mean.

• A policy aimed at a certain state of affairs — equilibrium deliberately pursued by a foreign office.
• An actual state of affairs — the distribution as it stands, whether or not anyone aimed at it.
• An approximately equal distribution of power — the strict equilibrium sense.
• Any distribution of power whatever — the loose sense, which he regarded as least defensible.

Morgenthau nonetheless treated the balance as a necessity rather than an option: given a plurality of states each seeking to preserve or upset the status quo, the configuration follows from the structure of the situation rather than from anyone’s preferences. See Realist Approach to International Relations.

Haas, Wight and Claude

• Ernst Haas identified eight distinct meanings in the literature — distribution, equilibrium, hegemony, stability and peace, instability and war, power politics generally, a universal law of history, and a prescriptive system for policymakers.

“Philological, semantic, and theoretical confusion.” — Ernst Haas, on the literature of the balance of power

• Martin Wight counted nine, among them the even distribution of power, the principle that it ought to be evenly distributed, equal aggrandisement among great powers, the principle that one’s own side ought to have a margin, the special role of a holder of the balance, and predominance itself.
• Inis Claude produced the cleanest taxonomy, and the one this article follows in substance.
  • Balance of power as a situation — an objective condition of the system.
  • As a policy — the conscious manipulation of that condition.
  • As a system — a whole pattern of international relations organised around such manipulation, with its own rules and roles.

Hume: common sense with an ancient pedigree

David Hume, in his essay on the balance of power, made the argument that has done most to keep the concept alive: the balance is not a modern European invention but a recurring practical intuition, visible wherever independent political units have coexisted.

• His historical claim is that Greek and Roman writers reasoned about coalitions against preponderant power in the terms eighteenth-century Europe used, though they lacked the phrase.
• His analytical claim is that the balance is not a theory requiring proof but ordinary prudence applied to politics, which is why it reappears in unrelated states systems.
• The reply to every announcement of the concept’s death: so long as common sense remains relevant, the balance of power will remain relevant.

Separate the description from the policy from the prediction, and most of the confusion turns out to be a quarrel between people answering different questions.

Origins and Evolution

The ancient Indian conception: Kautilya’s mandala

• Kautilya’s Arthashastra contains a fully developed indigenous balance-of-power theory, predating the European formulation by roughly two millennia. It is a theory of structural position, not of intention.
• The mandala, or circle of states, is its core device. The vijigishu, the would-be conqueror, sits at the centre and the rest are classified by geometric relation to him.
  • The immediate neighbour is the natural enemy (ari), because contiguity generates friction over territory, routes and succession; the neighbour’s neighbour is the natural friend (mitra), sharing the enemy without being adjacent, and the pattern alternates outward in rings.
• The six-fold policy (shadgunya) is the operational half — sandhi (peace), vigraha (hostility), asana (waiting), yana (marching), samshraya (sheltering with a stronger power, that is, alignment) and dvaidhibhava (peace with one while warring on another).
• The madhyama and udasina, the middle and indifferent kings, are Kautilya’s balancer — an external power strong enough to tip the outcome if it chooses.
• A structural theory of alignment arrived independently in South Asia, which strengthens Hume’s claim that balancing answers a general condition rather than expressing a European habit.

Greece and the Italian city-states

• Thucydides supplies the earliest narrative of a balance breaking down. The growth of Athenian power and the fear it caused in Sparta is the classic statement of preponderance producing a counter-coalition, and the Peloponnesian War is what followed when the balancing failed.
• The Italian city-state system of the fifteenth century produced the first self-conscious practice. Florence, Venice, Milan, Naples and the Papacy were of comparable size, close enough to threaten each other and too weak for any to conquer the peninsula — precisely the configuration the theory later specified.
  • Francesco Guicciardini described Lorenzo de’ Medici as deliberately keeping Italy in equilibrium by leaning to whichever side was weaker: the earliest explicit formulation of the balancer’s role.

Westphalia 1648: the container within which balancing became possible

• The Peace of Westphalia did not create the balance of power; it created the conditions under which balancing could become a system’s organising principle, and with them anarchy in the technical sense — absence of central authority, not disorder.
  • Sovereignty — supreme authority within a defined territory, with no superior.
  • Juridical equality of states irrespective of size, which makes combinations of the small against the large legitimate.
  • Non-intervention in others’ internal affairs, converting religion from a licence for war into a domestic matter.

Utrecht 1713: the phrase enters positive law

• The Treaty of Utrecht is the first treaty to write the balance of power into its own text, separating the French and Spanish crowns to secure the peace of Europe by an equilibrium of power.
• The war behind it is the model case of coalition balancing. The Grand Alliance against Louis XIV — England, the Dutch Republic, Austria, later Portugal and Savoy — was rebuilt across successive wars until the preponderance was contained, yet France remained a great power: containment without destruction is balancing rather than crusading.

Vienna 1815 and the Concert of Europe

• Napoleon’s empire was the second bid for continental hegemony, and its defeat is the second demonstration of the mechanism — and the standing example of its unreliability, since the coalitions formed slowly, broke repeatedly and took over twenty years to succeed.
• The Congress of Vienna, 1815, is where the balance stopped being a habit and became a managed institution.
  • Viscount Castlereagh sought a settlement restoring equilibrium rather than punishing France, returning it to the great powers so that it could counterweight Russia and Prussia.
  • Prince Metternich added the conservative layer — legitimate monarchical government, the Holy Alliance, and the containment of revolution as a source of systemic instability.
  • The mechanism was consultation: regular congresses of the five great powers, so that changes to the map were negotiated rather than seized. This is multilateral diplomacy in embryo.
• The Concert is credited with a hundred years without a general European war, 1815 to 1914 — with two qualifications: its congress phase weakened after the Crimean War of 1853–56, and the wars within the period were limited in aim.
• A.J.P. Taylor dated the golden age of the balance of power to 1848–1914, when the great powers manoeuvred continuously without any attempting to overthrow the system itself.

Bismarck’s system and its collapse into rigidity

• German unification in 1871 was the largest change to the European balance since Napoleon, because a state stronger than any other on the continent now sat at its centre — the outcome every previous settlement had been designed to prevent.
• Bismarck’s response was to manage what he had created through overlapping alignments — the Three Emperors’ League, the Dual Alliance with Austria-Hungary, the Triple Alliance with Italy, the Reinsurance Treaty with Russia — keeping Germany one of three in any grouping of five.
• After his dismissal the system froze. The Franco-Russian Dual Alliance of 1892, Britain’s Entente Cordiale of 1904 and the Anglo-Russian Convention of 1907 produced the Triple Entente facing the Triple Alliance — Britain joining the weaker grouping to check the strongest continental power, which is classic balancer behaviour.
• The critical change was rigidity of alignment, not the number of powers. Once alignment was fixed a Balkan quarrel could no longer be localised, and the mechanism designed to prevent general war became the belt that transmitted one.

The interwar failure and the Cold War’s bipolar version

• The interwar period tried collective security in place of balancing and got neither. The League failed over Manchuria and Abyssinia while balancing was disabled by American withdrawal, Soviet exclusion and Anglo-French unwillingness to act; appeasement followed. See Collective Security.
• The Cold War produced a balance with only two weights, which classical writers regarded as no balance at all. Alignment ceased to shift and the balancer role disappeared for want of a third great power.
• Nuclear weapons converted it into a balance of terror, resting not on comparable ability to win but on mutual inability to survive — a change of kind, treated in Deterrence.

Westphalia supplied the container, Utrecht the vocabulary, Vienna the machinery; 1914 supplied the warning that rigidity kills the mechanism.

Assumptions and Preconditions

The balance of power is a conditional law, not a universal one, and listing its conditions is the most economical way to assess its present relevance — because the case against relevance is precisely that the conditions have dissolved.

• Anarchy — no authority above states, so each provides its own protection. This produces self-help, and self-help is what balancing is.
• States are the primary actors, sovereign within their territories, and rational and survival-seeking — able to rank alternatives, estimate others’ power tolerably well, and prefer independence to submission.
• A plurality of significant actors, conventionally at least five. Two produce rigidity; three or four make a single defection decisive; five or more permit recombination.
• Roughly comparable units, of the same broad order of magnitude, so that combinations can actually offset each other.
• Flexibility of alignment — no permanent friends and no permanent enemies, only permanent interests, in the formulation attributed to Palmerston. Ideological blocs destroy this.
• A shared political culture — that the system is to be preserved, defeated powers readmitted, and victory not pressed to annihilation.
• Territorial divisibility — land or spheres available to compensate a power for another’s gain.
• A balancer — a power detached enough, and strong enough, to throw its weight to the weaker side without seeking dominion itself.
• Conventional warfare — war must be usable and proportionate, able to adjust the balance without destroying the participants — and capabilities must be observable well enough for states to know when a balance is tipping.

The Essential Rules of the Balance-of-Power System

Morton Kaplan turned the balance of power from a metaphor into a specified system by writing down the behaviour its members must follow and the structural condition it requires.

• The structural requirement is at least five essential national actors. Below that number the recombination the system depends on cannot occur, and the system transforms into some other type.
• The six essential rules, each an obligation on every essential actor.
  • Act to increase capabilities, but negotiate rather than fight. Power is to be accumulated cheaply; war is a residual instrument.
  • Fight rather than pass up an opportunity to increase capabilities. An actor that will not fight when the odds favour it will be discounted and eventually eliminated.
  • Stop fighting rather than eliminate an essential national actor. Victory is limited, because destroying a great power removes a future counterweight — the principle behind restoring France in 1815.
  • Act to oppose any coalition or single actor tending toward predominance. The anti-hegemonial rule, and the core of the system.
  • Act to constrain actors who subscribe to supranational organising principles. Universalist claims — imperial, revolutionary or ideological — threaten the plural structure itself and are opposed regardless of the claimant’s strength.
  • Permit defeated essential actors to re-enter the system as acceptable role partners, and treat all essential actors as acceptable partners. Today’s enemy must be available as tomorrow’s ally.
• The rules are not consistent in every instance, which is Kaplan’s own point: the second pushes toward war and the third pulls back, and stability depends on holding both.
• The critique is that they describe nineteenth-century Europe and little else. Of the six system types in Kaplan’s original set only the balance-of-power and loose bipolar models have ever existed — the four he added later are refinements of Cold War bipolarity rather than new departures — which makes the framework a classification rather than an explanation. His models as a whole are developed in Systems Theory and Scientific Approaches to International Relations.

Mechanisms and Techniques

The balance is not self-executing. The first division among its instruments is between building your own strength and borrowing someone else’s.

Internal and external balancing

• Internal balancing is capability built from a state’s own resources, and the more reliable of the two, because it depends on nobody else’s intentions.
  • Armament — force levels, modernisation, conscription and reserve depth — and economic development, since military capacity ultimately rests on the size and productivity of the economy, which makes growth a balancing act even when not intended as one.
  • Its cost is domestic and its advantage is autonomy — capability that is your own cannot be withdrawn at the moment of crisis.
• External balancing is the aggregation of others’ capability through alliance, alignment or partnership, which is faster and cheaper.
  • Its cost is dependence — an ally’s commitment is a promise about a future in which its interests may have changed.
  • Realists hold that states prefer internal balancing where they can afford it, resorting to alliance when a threat is too large or too near to be met alone.

The classical techniques

Alliances, the primary instrument, are treated in the next section. What follow are the remaining techniques, each a distinct way of moving weight between the pans.

• The balancer, or holder of the balance — a state that stays outside the main groupings and adds its weight to whichever side is weaker.
  • Britain played the role for roughly two centuries, aided by insular geography, naval rather than continental primacy, and an interest in a divided continent.
  • The role requires considerable strength, detachment from the quarrels being balanced, freedom from ideological commitment, and a reputation for intervening — a balancer who never acts stops being counted.
• Compensation and territorial adjustment — when one power gains, others receive equivalent gains so that relative standing is unchanged. Divide and rule, keeping a rival or region fragmented, serves the same end: a weak and divided Central Europe was for two centuries an object of French policy.
  • The partitions of Poland (1772, 1793, 1795) are the notorious case: Russia, Prussia and Austria preserved the balance among themselves by dividing a third state out of existence — the technique’s moral indictment in a single example.
  • Colonial compensation worked identically, African and Asian territory settling European scores. The technique has largely lapsed, the norm against conquest having removed the currency it traded in.
• Buffer states and buffer zones — a weak or neutralised state interposed between two strong ones, reducing friction by removing contact.
  • Belgium’s neutrality, guaranteed in 1839 to keep France and the German states apart, and Afghanistan, preserved as independent but weak between British India and Russian Central Asia, are the classic cases.
  • Nepal and Bhutan occupy the same position between India and China, which is why Chinese infrastructure and political penetration there registers in Delhi as a balance-of-power question rather than a bilateral one.
• Spheres of influence — an agreed or tacit division of regions in which each great power’s primacy goes uncontested, as in the Anglo-Russian Convention of 1907 over Persia.
  • The technique reduces friction between great powers at the cost of the states inside the spheres, which is why India resists the language of spheres even while acting on regional primacy.
• Armaments and arms races — competitive accumulation, which is internal balancing observed from outside.
  • The Anglo-German naval race before 1914 shows the logic and the pathology: each increment was defensive in intention and threatening in effect, which is the security dilemma working as the motor of balancing. See Security.
• Disarmament and arms control as balancing by agreement — capping capability by negotiation achieves equilibrium more cheaply than racing for it.
  • The Washington Naval Treaty of 1922 fixed capital-ship ratios among the major navies; Cold War strategic arms agreements did the same for warheads and launchers.
  • This is balancing, not its opposite — the parties agree where the equilibrium lies rather than renouncing it. See Deterrence.
• Intervention and non-intervention — the use, or withholding, of force in a third state’s affairs to prevent a rival’s gain: intervention adds a weight to a local contest, non-intervention denies a rival the pretext.
• War as the last resort of balancing — when no combination or adjustment will contain a rising power, war is the terminal instrument.
  • This is the concept’s deepest embarrassment: a device advertised as a preserver of peace counts war among its techniques, and the coalitions against Philip II, Louis XIV, Napoleon and Hitler were all wars to restore a balance.
  • Kaplan’s third rule marks the limit — stop fighting rather than eliminate an essential actor — which distinguishes balancing wars from wars of annihilation.
• Reparations and post-conflict settlements — the terms imposed after a war fix the distribution for the next period, so the choice between restoring and crippling the defeated power is itself a balancing decision.
  • 1815 restored France; 1919 crippled Germany, and the contrast is the standard illustration that removing a great power from the board destabilises, leaving a resentful revisionist with latent capacity to return.

Technique	What it moves	Standing example	Current viability
Alliance	Others’ capability into your pan	Triple Entente; NATO	High
Balancer	A detached weight to the lighter side	Britain in Europe	Low — requires multipolarity
Compensation	Territory as counterweight	Partitions of Poland	Effectively extinct
Buffer state	Distance between rivals	Belgium; Afghanistan	Surviving in the Himalayas
Spheres of influence	Agreed non-competition	Anglo-Russian Convention, 1907	Contested but practised
Arms racing	Own capability, competitively	Anglo-German naval race	High
Arms control	Capability downward, by agreement	Washington Naval Treaty, 1922	Weakening
War	The distribution itself	Coalitions against Napoleon	The residual instrument

Alliances

An alliance is the principal instrument of external balancing, and precision about what one is explains most of what happens to alliances.

What an alliance is, and what it is not

• An alliance is a formal association of states for the use, or non-use, of military force in specified circumstances against states outside its membership. Three elements do the work: it is formal, it concerns force, and it identifies contingencies. Because the commitment is prospective, credibility rather than capability is its scarcest resource.

Form	Formality	Content	Example
Alliance	Ratified treaty	Obligation to use force in defined contingencies	NATO, Article 5
Alignment	Informal, behavioural	Consistent expectation of support without a treaty	India–USSR through the Cold War
Coalition	Ad hoc	A single operation, dissolving on completion	The 1991 Gulf War coalition
Entente	Understanding	Shared outlook and consultation, no obligation to fight	Entente Cordiale, 1904
Alliance of convenience	Any	A shared enemy without shared values, expiring with the threat	The wartime Grand Alliance with the USSR

Why states ally

• Aggregation of capability — two weak states combining offset one strong one faster than either can grow.
• The autonomy–security trade-off — alliance buys security by spending independence, which is the axis on which India’s non-alignment was chosen.
• Restraining an ally — binding a dangerous partner inside a treaty to have a voice in its decisions, part of the reason for integrating West Germany into NATO.
• Signalling and deterrence — an alliance tells an adversary that an attack will meet more than one opponent, and access and infrastructure (bases, overflight, logistics, intelligence-sharing) is often the operative content even where the defence clause is never invoked.

Walt: states balance against threats, not against power

Stephen Walt’s balance-of-threat theory is the most important amendment the concept has received, because it repairs a fact the original gets wrong.

• The problem it solves: balance-of-power theory predicts combination against the strongest state, yet after 1945 most of Europe and Asia allied with the strongest state.
• Walt’s answer is that states balance against the greatest perceived threat, a compound of four elements rather than aggregate power alone.
  • Aggregate power, and geographic proximity — nearer power is more threatening, which is why Europe feared a Soviet Union on its border more than a United States across an ocean.
  • Offensive capability — the share of capability usable for attack, which makes posture and doctrine matter as much as force size.
  • Perceived intentions — a state believed to have aggressive aims is balanced against by states that would ignore an equally powerful but reassuring one.
• The gain is large. Balance of threat explains alliances the original calls anomalous and generates a prediction realism otherwise cannot make: a rising power can reduce the balancing against it by changing how it is read.
• The cost is precision — intentions are not measurable in advance, so the theory explains better after the event than before it.

Snyder’s alliance security dilemma

Glenn Snyder identified the bind inside every alliance: the same commitment that provides security creates two opposite fears, and reducing one increases the other.

• Abandonment — the fear that the ally will not honour the commitment when invoked: that it will realign, de-align, or simply fail to deliver.
• Entrapment — the fear of being dragged into a conflict over an ally’s interests one does not share, which is what alliance linkage did in 1914.
• The remedies are mutually defeating. Reassuring an ally reduces its fear of abandonment and increases your exposure to entrapment; hedging reduces entrapment and invites the ally to doubt you, and perhaps to seek another patron.
• The dilemma is sharper in multipolarity, where alternative partners exist; in a two-bloc system allies have nowhere to go, which is why Cold War alliances tolerated so much friction.

Burden-sharing and free-riding

• Alliance security is a collective good, protecting every member whether or not it paid, so each has an incentive to under-contribute.
• Olson and Zeckhauser’s finding — that in an alliance of unequal members the largest bears a disproportionate share, because it values the good most and will supply it alone — explains the history of American complaints about European defence spending.
• The contemporary form: NATO’s 2014 Wales pledge of 2% of GDP was replaced at the Hague summit of 25 June 2025 with 5% of GDP by 2035 — at least 3.5% core defence plus up to 1.5% defence-related investment, reviewed in 2029. European military spending rose 14% in 2025 to $864 billion.

Why alliances dissolve

• The threat disappears — the commonest cause, and what happened to the Warsaw Pact. Or members define it differently, as in CENTO and SEATO, whose members disagreed about whether communism, India or each other was the problem. Or the patron’s credibility erodes and members seek independent capability instead.

Alliances in the nuclear age

The claim that nuclear weapons made alliances obsolete and the claim that they made alliances indispensable are both defensible; the resolution is that they changed what alliances are for.

• The case that nuclear weapons reduced alliance value.
  • Capability aggregation stopped mattering at the top of the ladder. A secure second-strike force deters regardless of how many allies stand beside it.
  • Extended deterrence is inherently less credible than direct deterrence. Would the United States trade New York for Bonn? The question has no reassuring answer, since the guarantor is asked to accept its own destruction for a third party’s territory, and De Gaulle concluded that no state would — the basis of the force de frappe.
• The case that they made alliances more important.
  • Extended deterrence became the central product alliances sell. The American umbrella is why Germany, Japan and South Korea have not built arsenals, making the alliance system a load-bearing element of the non-proliferation regime.
  • Because the guarantee is doubted it must be continuously demonstrated — forward deployment, nuclear sharing, joint planning — producing denser integration than pre-nuclear alliances had.
  • Nuclear weapons are unusable for most purposes, so conventional balancing continues underneath them and still requires allies; and the stability-instability paradox means stability at the nuclear level permits more sub-nuclear conflict.
• The synthesis: alliances in the nuclear age aggregate credibility rather than capability, and their problem is reassurance rather than arithmetic.

The contemporary map

• NATO is the only fully functioning multilateral alliance of the classical type — 32 members after Finland in 2023 and Sweden in March 2024, an integrated command, and an Article 5 clause invoked once, after 9/11. Its enlargement in response to the war in Ukraine is the clearest recent instance of external balancing.
• The American system in Asia is hub-and-spoke rather than a bloc — separate bilateral treaties with Japan, South Korea, Australia, the Philippines and Thailand, no obligation running between the spokes, which maximises Washington’s leverage and minimises entrapment.
• AUKUS is a capability partnership rather than a mutual-defence treaty.
  • Pillar One covers nuclear-powered submarines for Australia: rotational US and UK deployments from Western Australia, sale of three to five Virginia-class boats, and a jointly built SSN-AUKUS design. A US review begun in June 2025 concluded in December 2025 with the agreement reaffirmed.
  • Pillar Two covers advanced capabilities — undersea warfare, hypersonics, quantum, AI and autonomy, cyber and electronic warfare — and its first signature project, announced 30 May 2026, is joint development of payloads for uncrewed undersea vehicles, in service from 2027.
• Minilateral groupings proliferate where formal alliances are unavailable — the Quad, the “Squad” grouping the United States, Japan, Australia and the Philippines on maritime security, and a lattice of trilaterals — delivering coordination without treaty obligation, which makes them acceptable to states unwilling to ally.
• The Quad is explicitly not an alliance, and India’s insistence on this is the price of its participation. Its 11th Foreign Ministers’ Meeting in New Delhi, 26 May 2026, operationalised the Indo-Pacific Partnership for Maritime Domain Awareness and launched a critical-minerals framework with a $20 billion target — public goods rather than collective defence.
• India has no alliances and one near-exception. The Indo-Soviet Treaty of August 1971, whose Article IX obliged each party to consult and to abstain from assisting a third party attacking the other, was a security guarantee in all but name.

Balancing and Its Alternatives

Balancing is one response to a rising power, not the only one. The alternatives make the theory testable, because a theory that calls every behaviour balancing predicts nothing.

Balancing against bandwagoning

• Bandwagoning is joining the stronger side rather than opposing it, buying safety by aligning with the source of the threat.
• Waltz’s case that balancing is the norm rests on what states most want: if survival comes first, joining the stronger side places your fate with a power you cannot restrain, while the weaker side makes you its more valued member.
• It supports him much less among small states, which bandwagon because they cannot affect the outcome and would only be punished for resisting. Walt observes that bandwagoning rises as a state’s own weight falls and the prospect of rescue recedes.
• Randall Schweller’s objection is that the theory carries a status-quo bias, assuming all states are security-seekers, which makes bandwagoning irrational by definition.
  • His correction distinguishes states by appetite — status-quo powers that want to keep what they have, and revisionist powers that want more. Bandwagoning for profit is the revisionist’s strategy: joining the winning side not from fear but to share the gains, as Italy and then the Soviet Union did in 1939–40.
  • The consequence is serious: if some states join aggressors opportunistically, counter-coalitions form more slowly than the model predicts, which is why balancing failed against Hitler.

Soft balancing and asymmetrical balancing

• Soft balancing — associated with Robert Pape and T.V. Paul — uses non-military instruments to constrain, delay or raise the cost of a dominant power’s unilateral action, on the logic that hard balancing against a preponderant power is unaffordable and provocative.
  • Brooks and Wohlforth object that it is indistinguishable from ordinary diplomatic friction, so calling it balancing makes the theory unfalsifiable.
  • Its instruments are institutional and diplomatic: withholding legitimacy in international organisations, coordinating positions, building rival institutions such as BRICS, the SCO, the New Development Bank and the AIIB, economic statecraft, and denying basing rights. The Franco-German-Russian coordination in the Security Council before the 2003 Iraq war is the type case.
• Asymmetrical balancing addresses threats from non-state actors, which have no territory to hold, no capability to count and no leadership to deter, and are met by intelligence, financial interdiction, law enforcement, special operations and denial of sanctuary.
  • A state cannot balance against an actor with no matching pan of the scales, which makes non-state threats a challenge to the concept rather than an application of it.

Hedging

• Hedging is the deliberate refusal to choose — deepening trade with a rising power and declining to join any alliance against it, while building the capability and partnerships that would be needed if engagement fails, and keeping both tracks below the threshold that would force a choice.
• It is the characteristic strategy of middle powers, and of most of Southeast Asia, because containment and accommodation are both too costly.
• India’s multi-alignment is hedging raised to doctrine — Quad membership alongside SCO and BRICS membership, American logistics agreements alongside Russian arms purchases.
• Hedging works only while both options stay open. A sharpening confrontation narrows the space, and the hedger’s advantage becomes the hedger’s exposure.

Buck-passing, chain-ganging and the other evasions

John Mearsheimer catalogued the strategies by which states avoid paying the costs of balancing themselves.

• Buck-passing — letting another threatened state bear the burden while you stay out and conserve strength, as Britain, France and the Soviet Union did to each other in the 1930s.
• Chain-ganging — the opposite failure, in which allies are so tightly bound that each is dragged into the others’ quarrels. This is 1914, and the mechanism behind the charge that the balance of power manufactures world wars.
• Bait and bleed — manoeuvring two rivals into a protracted war with each other while remaining aloof, so that both are weakened; bloodletting is the variant in which one belligerent is already your adversary and the object is to keep the war costly rather than end it.
• Balancing, buck-passing and bandwagoning are the three possible responses, and the theory cannot say in advance which will be chosen — though buck-passing is common in multipolarity, where responsibility is diffuse, and rare in bipolarity.
• Appeasement concedes substantive demands to a dissatisfied power to avoid conflict. It is not intrinsically foolish — it is correct toward a genuine status-quo power with legitimate grievances — but it fails catastrophically against a revisionist power with unlimited aims, which is what 1938 established.
• Neutrality is the refusal to take sides, and permanent neutrality depends on belligerents finding a neutral more useful than a conquest — Finland and Sweden joining NATO shows what happens when that changes. Non-alignment is distinct, being an active claim to judge each issue on its merits rather than a legal abstention.

Strategic engagement

Strategic engagement is the deliberate use of contact rather than confrontation to shape a rising power’s behaviour, and it is the most important non-balancing strategy available to a status-quo power.

• The concept: rather than contain a rising power, draw it into the existing order — trade, investment, institutional membership, dialogue, military-to-military contact — so that it acquires a stake in the order’s continuation and moderates its revisionism from inside.

Strategy	Assumption about the rising power	Instruments	Risk
Containment	Aims unlimited; only force restrains it	Alliances, forward deployment, technology denial	Provokes the confrontation it anticipates
Engagement	Aims negotiable and shaped by incentives	Trade, investment, institutional inclusion, dialogue	Funds the rise it hoped to moderate
Congagement	Aims undetermined, to be decided by circumstances	Both, simultaneously and deliberately	Mixed signals; each track undermines the other

• The application to China defines the concept. From the 1970s the United States and its partners engaged China comprehensively — normalisation, most-favoured-nation status, WTO accession in 2001, integration into supply chains — expecting a moderate stakeholder to result.
• The verdict has hardened. Engagement delivered the growth but not the moderation its advocates predicted, and the turn after 2017 to technology controls, supply-chain diversification and declared strategic competition is a judgement that engagement alone failed.
• The fairer assessment is that engagement was never pursued alone. It always ran alongside alliance maintenance and the pivot to Asia, which is what congagement was coined to name; the argument is about the mix, not about whether engagement was tried.
• India practises its own version. Trade with China reached about $155.6 billion in 2025 and the boundary is managed through the Special Representatives mechanism, whose 25th round met in 2026 — while India builds border infrastructure, raises defence spending and deepens the Quad. Engagement and balancing run on parallel tracks.

Polarity and the Stability of the System

Polarity is the number of great powers in the system, and the oldest quantitative question in the field is which number produces the fewest wars. The three configurations are defined in Power in International Relations; what follows is the argument about their consequences.

Unipolarity

• A unipolar system has one state whose capabilities no combination of others can match, so balancing against it is structurally unavailable.
• The case that it is stable, argued by William Wohlforth: the pole’s margin is so large that no rival can hope to close it, which removes both hegemonic war and great-power competition. Charles Krauthammer’s “unipolar moment” claimed the configuration would last longer than expected.
• The case that it is the least stable configuration is the stronger one.
  • The hegemon over-extends. Commitments accumulate faster than the capacity to fund them and resources consumed abroad are subtracted from the economic base that generated the primacy — imperial overstretch, in Paul Kennedy’s phrase — while free riders enjoy security and open markets without contributing.
  • Megalomania: with no external check, the dominant power comes to believe its power has no limits and begins acting unilaterally against the rules and institutions it built.
  • The standing examples are the 2003 invasion of Iraq without Security Council authorisation, which damaged the order’s legitimacy more than it damaged Iraq, and the WTO, whose Appellate Body has been unstaffed since December 2019 after an American refusal to appoint members — disabling dispute settlement in an institution the United States designed.
  • Balancing eventually returns, because unipolarity is precisely the condition Kaplan’s fourth rule and Waltz’s theory say the system will not tolerate.

Bipolarity

• Waltz’s case that bipolarity is the most stable configuration is the most influential single claim in the literature.
  • Fewer dyads, therefore less occasion for conflict — two great powers make one relationship to manage, five make ten.
  • No dependence on allies. Each superpower balances internally through its own armament, which removes entrapment and abandonment as system-level dangers.
  • Less miscalculation, because each side knows exactly who the adversary is; and each superpower contains anarchy in its own zone, disciplining allies and suppressing quarrels that would otherwise escalate, so much of world politics is hierarchically managed rather than anarchic.
• The counter-case.
  • Every conflict anywhere becomes systemically significant, because a gain for one side is a loss for the other; peripheral disputes in Korea, Vietnam, Angola and Afghanistan were fought as central ones. The relationship is also rigid — no third party to mediate, no balancer to correct an imbalance, and unbounded arms racing.
  • Cold War stability may be owed to nuclear weapons rather than to the number of poles, which would make the historical evidence inconclusive.

Multipolarity

• The case for multipolar stability turns on flexibility and on the arithmetic of attention.
  • Alignments can shift, so a rising power can be checked by recombination rather than by war.
  • Karl Deutsch and J. David Singer argued that as the number of independent actors rises, interaction opportunities rise faster, so the attention any state devotes to any one rival falls — diluting antagonism and making sustained arms racing against one target harder.
• The case against.
  • Greater uncertainty. With many powers it is harder to know who will fight alongside whom, and deterrence fails when the deterrer’s identity is unclear.
  • More fingers on the trigger — more states able to initiate a war that others are drawn into, and harder monitoring, so imbalances are detected late.
• Mearsheimer’s refinement is the most useful resolution, because it distinguishes two multipolarities the debate usually conflates.
  • Balanced multipolarity — several great powers of comparable strength, none with a realistic prospect of dominating. Relatively stable, though less so than bipolarity.
  • Unbalanced multipolarity — several great powers, one of which is a potential hegemon with a marked capability advantage. This is the most war-prone configuration of all, because the potential hegemon is tempted and the others are frightened, while buck-passing among the frightened delays the counter-coalition.
  • Pre-1914 and pre-1939 Europe are the two cases: Germany was a potential hegemon in both, and both produced world wars.

Configuration	Argument for stability	Argument against	Standing case
Unipolarity	No peer competitor, so no hegemonic war (Wohlforth)	Overstretch, free riding, megalomania and unilateralism	Post-1991 United States
Bipolarity	Fewer dyads, less miscalculation, internal balancing (Waltz)	Every periphery becomes central; total rigidity	The Cold War
Balanced multipolarity	Flexible alignment, diluted attention (Deutsch and Singer)	Uncertainty, misperception, buck-passing	The Concert of Europe
Unbalanced multipolarity	None claimed	The most war-prone configuration (Mearsheimer)	Europe, 1914 and 1939

The benefits of a multipolar world

Multipolarity is defended on grounds other than the frequency of war, and those grounds are why most states prefer it.

• Strategic autonomy for middle and smaller powers, who can bargain among competing great powers rather than accept the terms of one. India’s ability to buy Russian air defence while deepening American defence ties exists only because the system has more than one pole.
• Restraint on unilateralism, since multiple centres mean multiple vetoes, so no single state can act outside agreed rules without cost.
• Institutional choice and civilisational plurality — competing development finance, technology stacks and trade arrangements give developing states alternatives, and no single model of development is presented as the only legitimate one.
• Democratisation of global governance, since multipolarity supports reform of institutions whose authority still reflects 1945 — the Security Council’s permanent membership, the Bretton Woods quotas — and creates the coalitions able to press it, as India does through the G4.
• The honest qualification: these are benefits of distributed authority rather than of multipolarity as such, and they coexist with the higher risk of miscalculation. A multipolar world is more equitable and less predictable at once.

The states that most want a multipolar world want it for the autonomy it gives them, not for a peace it might not deliver.

Is polarity still a useful frame?

• The count depends on which power is measured, and different measures give different answers.
  • The same system is unipolar militarily — American defence spending of $954 billion in 2025 against China’s $336 billion — bipolar economically, since the United States and China dominate the technology and trade frontier, and multipolar diplomatically. A frame yielding three answers for one system is not describing a single property.
• Joseph Nye’s three-dimensional chessboard makes the same point about issue-specific power: the top board of military power is close to unipolar, the middle board of economic power has been multipolar for decades, and the bottom board of transnational relations — climate, pandemics, terrorism, contagion, cyber — has no polarity at all, power there being distributed among states, firms and networks.
• Non-state actors sit outside the frame entirely. Firms whose revenues exceed most states’ budgets exercise consequential power without being poles: the 2026 Fortune Global 500 reported combined revenues of $43.1 trillion. See Transnational Actors.
• Richard Haass’s “non-polarity” is the sharpest form of the objection: what defines the present system is not the number of poles but the number of consequential actors — states, regions, firms, NGOs, militias, cartels — none dominant.
• Amitav Acharya’s “multiplex world” is the most useful reformulation, because it replaces the frame rather than merely criticising it.
  • The image is a multiplex cinema — several films running in adjacent halls with different producers and plots, rather than one global blockbuster.
  • The claims: the era of a single hegemon or single ordering principle is over; regional orders are increasingly self-sustaining; the Global South produces order rather than consuming it; and interdependence persists even as leadership fragments, so the outcome is decentralised order rather than disorder.
  • Acharya has continued to develop the argument, contending most recently that the decline of Western primacy is not the decline of global civilisation, and that an order can be plural in its sources of authority without collapsing.

Theoretical Interpretations

The realist reading

• For structural realists the balance is a consequence, not a policy. Anarchy plus units that wish to survive is sufficient to produce balancing; no shared culture, diplomatic tradition or intention is required. The cost of so strong a claim is that the theory predicts the outcome without predicting any state’s behaviour.

“Balance-of-power politics prevail wherever two, and only two, requirements are met: that the order be anarchic and that it be populated by units wishing to survive.” — Kenneth Waltz

• Mearsheimer’s offensive variant denies that states balance to preserve equilibrium at all: they maximise relative power and seek regional hegemony, and balancing is what other states do to them — unreliably, since buck-passing is often cheaper.

The liberal and idealist critique

• Woodrow Wilson’s indictment is the foundational one: the balance of power does not prevent war but produces it, because competitive alignment converts every quarrel into a general one — and it is indifferent to who is right, treats small states as counters, and rests on secret diplomacy and arms racing.
• The measles-and-plague formulation captures it — protection against measles while creating the conditions for plague, a local dispute being transmitted through alliance linkage into a world war. 1914 is the demonstration.
• Nehru’s objection was different in kind. He described the balance as producing a “nervous state of peace” — not peace but the suspension of war under continuous fear, sustained by armament and the permanent expectation of conflict. The critique concerns the quality of the peace, not its duration.

The constructivist critique

• The material structure is not self-interpreting. Whether a build-up counts as a threat depends on shared understandings about who is building, so a rising power is balanced against when identities and norms make it legible as a threat.
• Anarchy is what states make of it, in Alexander Wendt’s formulation — anarchy has no single logic, and self-help is one culture among several rather than a necessary consequence of the absence of government.
• The decisive illustration: British nuclear weapons do not threaten the United States and North Korean ones do, though the British arsenal is far larger, and no purely material theory produces that result.

The English School: the balance as an institution

• The English School treats the balance of power as one of the institutions of international society,, alongside diplomacy, international law, war and the great powers — a shared practice sustained by common understandings rather than a mechanical outcome.
• Hedley Bull distinguishes a fortuitous balance, arising by accident, from a contrived balance, maintained deliberately because states believe it should be. So understood, the balance preserves the states system against universal empire and protects the independence of small states and regional subsystems.

Alternatives and Challengers to the Concept

Collective security

Collective security is the institutionalised alternative to balancing and its most important rival. The two are usually presented as opposites, but share more than the contrast suggests — which is why the relationship is called complementary and antagonistic at once.

• What they share: both are methods of managing power rather than of abolishing it; both rest on the proposition that greater power is the antidote to power; and both treat sovereignty as sacred while accepting war as a legitimate instrument for defending it.
• Collective security is therefore an institutionalised form of the balance of power, which is what makes the two complementary rather than simply opposed.

	Balance of power	Collective security
Operating condition	Anarchy; no organisation required	Requires a functioning international organisation
Character	Ad hoc and improvised	Institutionalised and prior
Enemy	Unknown in advance; anyone may become one	Unknown, but action is triggered by defined aggression
Certainty	Uncertain — a counter-coalition may or may not form	Theoretically certain — all are pledged in advance
Principal beneficiary	Great powers, who manipulate it	Small and weak states, who cannot balance alone
Effect on armament	Encourages arms racing and alliance-building	Intended to reduce both
Standing weakness	Produces the wars it claims to prevent	Has never been reliably operationalised

• Uncertainty is the balance of power’s real weakness, and Napoleon is the illustration: when a state set about destroying the European order, no counter-coalition formed automatically and it took over two decades before one succeeded.
• Neither has displaced the other. Collective security has never been operationalised reliably under the League or the United Nations, so states have not abandoned alliances and power politics — but the balance of power has never dispensed with the legitimating language of collective action either. The machinery, the League and UN record, the veto and R2P belong to Collective Security.

Power transition theory, deterrence and security communities

• A.F.K. Organski’s power transition theory is a rival prediction rather than a rival prescription: the system is hierarchical rather than anarchic, and war is most likely not when power is unequal but when a dissatisfied challenger approaches parity with the dominant power.
  • This inverts the balance-of-power expectation. Balance theory says equality preserves peace and preponderance invites war; transition theory says the opposite, and both cannot be right. See Power in International Relations.
• Bernard Brodie’s claim is that nuclear weapons replace balancing rather than modify it: what a state needs is not an equilibrium of capability but an assured ability to retaliate, after which the adversary’s superiority becomes irrelevant — dissolving the arithmetic the tradition rests on.
  • The qualification is that nuclear weapons are unusable for almost everything, so conventional balancing continues below the nuclear threshold — which is why the concept survives. See Deterrence.
• A security community, in Karl Deutsch’s formulation, is a group of states among whom war has become unthinkable, and its existence in Western Europe removes that region from the game altogether.
  • Western Europe is the case. Between France and Germany, the axis of four wars in seventy years, balancing has ceased to be intelligible. The balance of power is therefore not a permanent condition of the states system but a feature of relationships of a particular kind, which can be transformed.

Relevance Today

The case that the concept has lost its conditions

The strongest argument against relevance is not that the balance of power is wrong but that the world it described has gone: each precondition has weakened.

• Politics has become global rather than regional, and the Westphalian order has been overlaid by dense institutions and complex interdependence. The balance is intelligible as a regional mechanism with identifiable neighbours; a genuinely global system has no manageable set of weights.
• The configuration has been wrong for most of the concept’s modern life. Balancing operates in multipolarity, and the last eighty years have been bipolar and then unipolar.
• Warfare has changed in kind. Nuclear weapons, cyber operations, precision strike and autonomous systems break the assumption that capability can be counted and used proportionately.
• Threats come from actors that cannot be balanced. Non-state actors operate from any soil, prefer failed states as sanctuary, and offer no target balancing logic can identify.
• The moral objection persists. A system that partitioned Poland and traded colonies as compensation has no answer to the charge that it is order purchased at the expense of the weak.

The case that the concept has returned

• Great-power competition has returned, and with it the plural distribution the concept requires — the rise of China, the resurgence of Russia and the emergence of India as an independent weight.
• The Asia-Pacific is a live balancing theatre, which is why the centre of gravity of world politics has shifted there: American forward deployment, the rebalance to Asia, and region-wide acquisition of conventional capability.
• China’s neighbours are aligning and quasi-aligning — the Quad, the Squad, AUKUS, upgraded American treaty relationships with Japan and the Philippines, Japanese rearmament — mostly without formal alliance.
• NATO’s enlargement and rearmament are the European instance. Two historically neutral states joined after a single act of aggression, and European military spending rose 14% in 2025 to $864 billion.
• Nuclear weapons have not abolished conventional balancing, because they are unusable for anything short of survival. World military expenditure reached $2,887 billion in 2025, 2.5% of world product — not the pattern of a world that has stopped balancing.
• The honest conclusion is that the balance of power has ceased to be the organising principle of the whole system and remains the organising principle of its most dangerous regions. That is a demotion, not an obituary.

The concept survives not because the world matches its model but because no rival has explained great-power behaviour better.

The Indian Experience

India is the most instructive single case in the concept’s contemporary application: it produced an indigenous theory of balancing, a sustained doctrinal rejection of it, and a return to its practice under other names.

Theory, rejection and the exception

• The mandala is India’s contribution to the theory rather than to its practice — alignment following from position, and enmity a property of adjacency rather than of intention. Its contemporary echo is unavoidable, since China is both the immediate neighbour to the north and the neighbour’s neighbour to the west.
• Nehru rejected the balance of power on moral and practical grounds — as a system that produced two world wars and treated newly independent states as counters, and as a mechanism that would subordinate India’s freshly won autonomy to someone else’s quarrel.
• The “nervous state of peace” formulation is the core of the objection: a balance sustains not peace but the deferral of war, at the price of continuous armament and continuous fear.
• Non-alignment was an attempt to escape the mechanism rather than to play it better — refusing bloc membership and expanding the room in which states outside the blocs could act.
• The realist reading is less flattering and probably more accurate: non-alignment maximised India’s bargaining position with both blocs and avoided the costs of alliance — a sophisticated hedging strategy conducted in the language of principle.
• The Indo-Soviet Treaty of August 1971 is the standing exception, concluded as the Bangladesh crisis moved toward war and the United States and China aligned with Pakistan. Its consultation clause deterred Chinese intervention — balancing under another name.

Balancing China after 1998 and after 2020

• The 1998 nuclear tests were justified by the Chinese capability and the China-Pakistan relationship, making them internal balancing at the strategic level; the Galwan clash of 2020 was the decisive turn, converting a managed boundary dispute into a standing confrontation.
• Internal balancing since: defence expenditure of $92.1 billion in 2025, up 8.9%, the world’s fifth-largest, accelerated Himalayan border infrastructure, and indigenous defence production.
• External balancing: the Quad, an Indo-Pacific maritime posture, and the foundational logistics and communications agreements with the United States that make interoperability possible without an alliance.
• Soft balancing: BRICS, whose eleven full members now include Egypt, Ethiopia, Iran, the UAE, Indonesia and Saudi Arabia (which has not formally confirmed accession), and the SCO — constraining Chinese and American unilateralism without confronting either.
• Engagement runs alongside all of it. Bilateral trade reached about $155.6 billion in 2025, direct flights resumed, and the Special Representatives mechanism held its 25th round in 2026. Disengagement at the friction points has been achieved; de-escalation and de-induction have not.

Balancer or swing state, and the tension India does not resolve

• The classical balancer needs strength, detachment and willingness to act, and India possesses the first two more clearly than the third. The swing-state reading is more accurate: India is large enough that its alignment materially changes the regional balance and unaligned enough that both sides must court it.
• A balancer must be detached from the quarrel it balances, and India is a direct party to the contest with China, so it cannot hold the balance between China and others.
• The two-front problem is the structural difficulty — a militarily superior adversary to the north and a hostile, nuclear-armed adversary to the west, with a deepening relationship between them. This is the classic case for external balancing, since a state facing two aligned adversaries cannot generate sufficient capability internally.
• Operation Sindoor, following the Pahalgam attack of 22 April 2025, and the “new normal” articulated afterwards — terrorism treated as an act of war, the Indus Waters Treaty held in abeyance — is the sub-conventional face of the same problem.
• India advocates a multipolar world globally and regional primacy in South Asia, and the two rest on incompatible premises. The global argument is that concentrated power invites unilateralism; the regional argument is that external powers should acquire no foothold in South Asia and Indian preferences should carry disproportionate weight — which is the hegemonic case, not the multipolar one.
• India’s smaller neighbours read the contradiction accurately, which is why Nepal, Sri Lanka, Bangladesh and the Maldives have used China as a counterweight — practising against India the balancing India practises against China.
• The tension is not hypocrisy so much as the ordinary condition of a rising power. It remains the sharpest test of whether India’s multipolar advocacy is a principle or a stage.