Skip to content

Market power

Gauge an actor's ability to profitably set price above (or wages below) the competitive level by reading the slope of the downward-sloping residual demand curve it faces, quantified as the price-cost wedge (P − MC)/P.

Core Idea

Market power is the ability of a seller — or, in the mirror case of monopsony, a buyer — to set a price profitably above (or below) the competitive level and sustain it without losing all sales (or all supply). The structural condition is a downward-sloping firm-level residual demand curve: buyers cannot costlessly substitute to rivals, so the firm faces a trade-off between price and volume rather than a perfectly elastic market that would punish any above-cost price with complete defection. The firm exercises market power by choosing the price–quantity pair along this residual demand curve that maximises profit, which under standard assumptions produces a price above marginal cost and a quantity below the socially efficient level.

The canonical diagnostic is the Lerner index — (P − MC) / P — which is zero under perfect competition and positive wherever market power exists, with its value equal to the reciprocal of the absolute price elasticity of demand facing the firm. A large firm in a competitive market with many close substitutes can have a Lerner index near zero; a small firm with a highly differentiated product or a captive customer base can have a Lerner index well above zero. The sources of market power in product markets include switching costs, search frictions, geographic isolation, network effects, intellectual property rights, scale economies that make profitable entry difficult, and strategic conduct designed to deter rivals. In labour markets (monopsony), the analogous sources are geographic immobility, job-specific skills, employer coordination, and the costs of job search, which allow employers to pay wages below the marginal revenue product of labour.

Market power generates two welfare-relevant consequences: a distributional one, in which surplus is transferred from buyers (or workers) to the powerful firm, and an efficiency one, in which units that would be mutually profitable at marginal-cost prices are not traded, producing a deadweight loss measured by the standard welfare triangle between the demand curve, the marginal-cost curve, and the monopoly price. The growing macro-level evidence from De Loecker, Eeckhout, and colleagues that aggregate markups in the US economy rose substantially after 1980 has linked market power to declining labour shares, reduced investment rates, and widening productivity dispersion — extending what was once a firm-level industrial-organisation concept into a macroeconomic concern.

Structural Signature

Sig role-phrases:

  • the firm (or buyer) facing a sloped residual curve — an actor confronting a finite, downward-sloping firm-level residual demand curve (or, in monopsony, an upward-sloping supply curve) rather than the flat curve of perfect competition
  • the substitution friction — what produces the slope: switching costs, search frictions, geographic isolation, network effects, intellectual property, scale-driven entry barriers, strategic conduct (job-specific skills and search costs in the labour mirror) preventing the other side from costlessly leaving
  • the price–quantity choice along the curve — the exercise of power: choosing the profit-maximising point, which yields price above marginal cost and quantity below the efficient level
  • the price–marginal-cost wedge (Lerner index) — the diagnostic scalar (P − MC)/P, zero under perfect competition and positive wherever power exists, equal to the reciprocal of the residual-demand elasticity
  • the size-independent test — power read off the slope, not market share: a small differentiated firm can be powerful and a large hemmed-in one powerless
  • the distributional consequence — surplus transferred from buyers (or workers) to the powerful side
  • the efficiency consequence (deadweight loss) — mutually profitable trades that do not happen at the above-cost price, recorded as the welfare triangle
  • the monopsony mirror — buyer power as the identical residual-curve structure read from the other side (wages suppressed below marginal revenue product)
  • the aggregation property — the Lerner scalar summing across firms, carrying the firm-level concept up to economy-wide markups, labour share, investment, and productivity dispersion

What It Is Not

  • Not bigness or market concentration. What matters is the slope of the firm's residual demand curve, not its share. A dominant-share firm hemmed in by close substitutes has little power (raise price, lose nearly all sales), while a small firm with a differentiated product or captive base can have real power. Reading market share as power conflates a large firm with a powerful one.
  • Not the same as having a positive accounting markup. Price above marginal cost is the diagnostic, but a markup can reflect recovery of large fixed costs or genuine scarcity rather than the ability to profitably restrict output. The structural condition is a downward-sloping residual demand curve the firm exploits, not any gap between price and marginal cost taken in isolation.
  • Not inherently illegal or abusive. Possessing market power is not by itself an antitrust violation; firms can acquire it through superior products, innovation, or scale. The welfare concern attaches to its exercise and the conduct that entrenches it, so "has market power" is a structural finding, not a verdict of wrongdoing.
  • Not its own diagnostic, the Lerner index. The index (P − MC)/P meters market power and equals the reciprocal of the residual-demand elasticity, but the power is the underlying capacity to set price above the competitive level; the scalar is the readout, not the thing. A measured markup is evidence of power, not identical to it.
  • Not bargaining power in general. Market power is the price-theoretic specialization — a residual demand curve, marginal cost, surplus accounting — whereas bargaining_power is the broader pattern wherever one side's outside option is worse. Cross-domain "platform power" or "narrative power" invoke that parent or bottleneck control, not the residual-demand-and-deadweight-loss apparatus that defines market power; importing the economic machinery there distorts.

Scope of Application

Market power lives across the markets studied by economics — product, labour, capital, and platform — wherever an actor faces a finite, downward-sloping residual response curve to its own price (or wage) choices; its reach is within that domain. The cross-substrate look-alikes ("platform power," "narrative power") rest on different structural commitments and travel under the parents bargaining_power and bottleneck, not under the residual-demand-and-deadweight-loss apparatus that defines market power here.

  • Industrial organization — the canonical product-market home; the Lerner index, Herfindahl-Hirschman concentration, residual-demand estimation, and the Cournot/Bertrand/Stackelberg oligopoly models all operationalize the slope of the firm's residual demand.
  • Antitrust enforcement — Sherman/Clayton, EU Article 102, and merger review keyed to HHI thresholds and diversion ratios, where market definition is the prelude to measuring whether a firm can profitably restrict output.
  • Labour economics (monopsony) — the buy-side mirror: a concentrated employer facing an upward-sloping labour-supply curve suppresses wages below the marginal revenue product of labour, the identical residual-curve structure read from the other side.
  • Macroeconomics — the rising-aggregate-markup literature (De Loecker–Eeckhout) sums the Lerner scalar across firms to carry a once firm-level concept up to labour share, investment, and productivity dispersion.
  • Regulation — rate-of-return and price-cap oversight of natural monopolies (utilities, common carriers), each of which presupposes a market-power measurement before constraining the firm's pricing.

Clarity

Naming market power separates a large firm from a powerful one — a distinction that raw size and concentration figures constantly run together. What matters is not how big the firm is but whether it faces a downward-sloping residual demand curve: a dominant-share firm hemmed in by close substitutes can raise price and lose essentially all its sales (no power), while a small firm with a differentiated product or a captive customer base can raise price and keep most of them (real power). The concept thereby reframes the loose enforcement question "is this firm too big?" into the sharper, answerable one "what is the slope of its residual demand curve, and what produces that slope?" — and that second question decomposes cleanly into the substitution frictions that generate power: switching costs, search frictions, geographic isolation, network effects, intellectual property, scale-driven entry barriers, strategic conduct. The diagnostic test follows directly: if the firm can raise price above cost without losing all custom, market power is present, whatever its market share.

The concept also keeps two welfare consequences from being conflated, which matters because the policy responses differ. There is a distributional effect — surplus transferred from buyers (or, in the monopsony mirror, from workers) to the powerful firm — and an efficiency effect — mutually profitable trades that simply do not happen at the above-cost price, recorded as the deadweight-loss triangle. Holding these apart lets the analyst see that a markup is not merely a transfer to be redistributed and not merely a deceased trade to be lamented, but both at once, and that the monopsony case is the same structure read from the buy side: an employer facing an upward-sloping labour-supply curve suppressing wages below marginal revenue product. That structural unification — product-market seller power and labour-market buyer power as one residual-curve phenomenon — is what lets the recent finding of economy-wide markup growth carry a once firm-level concept into questions of labour share, investment, and productivity dispersion.

Manages Complexity

The settings where a firm might extract surplus are wildly heterogeneous on their surface — a regional trauma centre bargaining with insurers, a search-ads platform, a domestic brewer, a single dominant employer in a company town — and each comes wrapped in its own institutions, contracts, technologies, and sources of advantage: switching costs here, network effects there, geographic isolation, intellectual property, scale-driven entry barriers, strategic entry deterrence, in the labour mirror job-specific skills and search costs. Market power compresses this entire zoo onto a single structural object, the firm-level residual demand curve, and a single scalar read off it, the Lerner index (P − MC)/P, equal to the reciprocal of the elasticity that curve embodies. Every disparate case becomes the same question — what is the slope of this firm's residual demand, and is the price-cost wedge positive? — so the analyst tracks one number rather than cataloguing each industry's idiosyncrasies, and the long list of power's sources is reorganised not as separate phenomena but as so many distinct generators of one quantity: the slope. The diagnostic then reads off that scalar directly, independent of market share, which is exactly what lets a small differentiated firm register as powerful and a large hemmed-in one as not.

The compression's reach is widened further by two unifications that fold still more of the domain into the same apparatus. First, the monopsony mirror: buyer power in a labour market — an employer facing an upward-sloping supply curve and suppressing wages below marginal revenue product — is not a separate concept requiring its own machinery but the identical residual-curve structure read from the other side, so product-market and labour-market power collapse into one object the analyst handles with one toolkit. Second, the macro extension: because the Lerner index is a single comparable scalar, it aggregates, and the finding that economy-wide markups rose after 1980 carries a once firm-level concept up to questions of labour share, investment, and productivity dispersion without any new structural primitive — the same number, summed. And the welfare reading attaches as a clean two-branch readout of that one object rather than a fresh analysis per case: any positive wedge simultaneously implies a distributional transfer of surplus to the powerful side and an efficiency loss recorded as the deadweight triangle of trades that do not happen. So from one curve and one scalar the analyst reads slope, source, presence-of-power, buy-side or sell-side, aggregate trend, and both welfare consequences — a high-dimensional landscape of markets and harms reduced to a residual-demand slope and the wedge it produces.

Abstract Reasoning

Market power licenses reasoning that routes every question about a firm's capacity to extract surplus through one structural object — the slope of its residual demand curve — and reads diagnosis, source, welfare, and policy off that slope and the price-cost wedge it produces.

The foundational move is the test for power that bypasses size. To judge whether a firm has market power, the analyst does not consult its market share but asks whether it can raise price above cost without losing all custom — equivalently, whether it faces a downward-sloping rather than a flat residual demand curve. The reasoning runs from the responsiveness of its sales to a price increase to the presence of power, so a dominant-share firm hemmed in by close substitutes registers as powerless (raise price, lose essentially all sales) while a small firm with a differentiated product or captive base registers as powerful (raise price, keep most). This is the move that reframes the loose "is this firm too big?" into the answerable "what is the slope of its residual demand, and is the price-cost wedge positive?"

The decisive measurement move is quantifying power as a scalar with a built-in dual reading. The analyst computes the Lerner index (P − MC)/P and reasons that, equal to the reciprocal of the absolute elasticity of the firm's residual demand, it simultaneously states how much power the firm has and how inelastic the demand it faces is. So reading the markup off cost data is reading the curve's slope at the same time, and the analyst tracks one number rather than cataloguing each industry's idiosyncrasies — the disparate cases collapse to the same question.

A third move is source-decomposition of the slope. Rather than treat each industry's advantage as sui generis, the analyst reasons that the long list of power's origins — switching costs, search frictions, geographic isolation, network effects, intellectual property, scale-driven entry barriers, strategic conduct, and in the labour mirror job-specific skills and search costs — are not separate phenomena but so many distinct generators of one quantity, the slope. This licenses the inference that asking "what produces this firm's power?" is asking "what makes its buyers unable to costlessly substitute?", so the analyst diagnoses the specific friction responsible and targets it, treating the residual-curve steepness as the common downstream variable every source feeds.

A fourth move is the two-branch welfare readout from a single wedge. Any positive price-cost wedge licenses two simultaneous inferences the analyst keeps distinct because the policy responses differ: a distributional effect (surplus transferred from buyers — or, in the monopsony mirror, workers — to the powerful side) and an efficiency effect (mutually profitable trades that do not happen at the above-cost price, recorded as the deadweight-loss triangle). The reasoning reads both consequences off the same object rather than performing a fresh analysis per case, and it warns the analyst that a markup is neither merely a transfer to redistribute nor merely a lost trade to lament but both at once.

A fifth move is mirror-symmetry to the buy side. The analyst reasons that monopsony is the identical residual-curve structure read from the other side — an employer facing an upward-sloping labour-supply curve and suppressing wages below marginal revenue product — so buyer power requires no new machinery and product-market and labour-market power are handled with one toolkit. The move is to recognize a sloped response curve to an actor's own choices, whichever side of the transaction the actor sits on, and apply the same slope-and-wedge reasoning.

Finally, the concept supports an aggregation-and-scale-up move. Because the Lerner index is a single comparable scalar, the analyst reasons that it sums across firms, so the finding that economy-wide markups rose after 1980 carries a once firm-level concept up to macro questions — declining labour shares, reduced investment, widening productivity dispersion — without any new structural primitive. The reasoning is that the same number, aggregated, licenses claims about the whole economy, letting a firm-level diagnostic become evidence about economy-wide surplus extraction and its macro consequences.

Knowledge Transfer

Within economics market power transfers as mechanism, because the structural object that defines it — an actor facing a finite, downward-sloping residual response curve to its own price (or wage) choices, behind which lies a substitution friction that prevents the other side from costlessly leaving — recurs across the field's markets with its machinery intact. The test for power that bypasses size, the Lerner-index quantification with its built-in elasticity dual, the source-decomposition of the curve's slope into specific frictions, and the two-branch welfare readout (distributional transfer plus deadweight-loss efficiency cost) all carry without translation across industrial organization (the canonical product-market home, with concentration indices, residual-demand estimation, and the Cournot/Bertrand/Stackelberg oligopoly models), antitrust enforcement (Sherman/Clayton, EU Article 102, merger review keyed to HHI and diversion ratios), labour economics (the monopsony mirror: an employer facing an upward-sloping labour-supply curve suppressing wages below marginal revenue product — the identical residual-curve structure read from the buy side, requiring no new machinery), regulation (rate-of-return and price-cap oversight of natural monopolies, which presuppose a market-power measurement), and the macro markup literature (De Loecker–Eeckhout: because the Lerner index is a single comparable scalar it aggregates, carrying a once firm-level concept up to labour share, investment, and productivity dispersion without a new primitive). Across all of these the apparatus — residual demand, marginal cost, the price-cost wedge, deadweight loss — travels as the working machinery, because product, labour, capital, and platform markets are co-instances of one sloped-response-curve structure. This is genuine within-domain mechanistic reach.

Beyond economics the transfer is best read as a shared abstract mechanism rather than the named concept traveling, and the boundary needs care because the surface cases look so similar. What genuinely recurs across substrates is the more general pattern market power instantiates: an actor faces a sloped — not flat — response curve to its choices because the alternatives available to the other side are imperfect substitutes, so it can extract surplus by moving along the curve. That pattern is carried at the prime level by bargaining_power (the broader, not-only-product-market version, where one side's outside option is worse than the other's) and by bottleneck / critical-path-resource control (an essential input conferring an analogous extractive position). Those parents travel cross-domain as genuine mechanism. But the cross-substrate analogues invoked as "platform power" or "narrative power" must be marked carefully: they are real, yet they rest on different structural commitments and are not all special cases of the same primitive — invoking "market power" for them imports specifically economic apparatus (the residual demand curve, marginal cost, the deadweight-loss triangle, surplus accounting) that does not translate without distortion, so for them the named concept travels only by analogy while the genuine mechanism is the more general bargaining-power or bottleneck pattern. Strip the price-theoretic apparatus and what remains is exactly the sloped-response-curve, surplus-extraction pattern of those parents, not market power. So the honest move when reaching cross-domain is to carry the parents (bargaining power, bottleneck control) and to leave "market power," as named, at home with its demand-curve-and-marginal-cost specifics, where product-market seller power and labour-market buyer power are genuine co-instances. (Its own diagnostic, the Lerner index, is the scalar that meters this same structure, and the family of nearby concepts — monopoly rent as the realized surplus, network effects and switching costs as sources, rent-seeking as the political-economy response — orbit the same object.) The boundary between the home-bound named concept and the traveling sloped-response-curve mechanism is drawn in full in Structural Core vs. Domain Accent.

Examples

Canonical

The defining measure is Abba Lerner's index (1934). Suppose a firm profit-maximises at a price of $10 per unit while its marginal cost of producing that unit is $6. The Lerner index is (P − MC)/P = (10 − 6)/10 = 0.4 — a positive price-cost wedge, so the firm has market power; under perfect competition price would equal marginal cost and the index would be 0. Because a profit-maximising firm sets its markup equal to the reciprocal of the elasticity it faces, an index of 0.4 implies the firm's residual demand has an absolute elasticity of 1/0.4 = 2.5: a 1% price rise costs it about 2.5% of sales. The same $10 price at a firm with marginal cost $9.80 gives an index of 0.02 and an elasticity of 50 — near-competitive, whatever either firm's market share.

Mapped back: (P − MC)/P is the price–marginal-cost wedge (Lerner index); its value pins down the slope of the curve the firm facing a sloped residual curve confronts (elasticity 2.5 versus 50). That the near-competitive reading can attach to any market share is precisely the size-independent test — power is read off the wedge, not bigness.

Applied / In Practice

Competition authorities operationalise this test in merger review through the "hypothetical monopolist," or SSNIP, test in the US Horizontal Merger Guidelines. Rather than ask whether a merged firm would be large, regulators ask whether a hypothetical sole supplier of the candidate products could profitably impose a Small but Significant Non-transitory Increase in Price — conventionally about 5%. If enough customers would defect to substitutes to make the increase unprofitable, the residual demand is too flat and the market is drawn wider; if they would not, the firm faces a downward-sloping residual demand and the power is real. Diversion ratios — what share of lost sales flow to which rival — estimate that slope directly. The same substitution logic now drives labour-market cases, where enforcers scrutinise no-poach agreements as devices that steepen employers' residual labour-supply curve.

Mapped back: The SSNIP test is the size-independent test made procedural: can the actor profitably raise price? A "yes" means the firm facing a sloped residual curve is real, and diversion ratios measure the substitution friction producing the slope. Extending it to no-poach cases is the monopsony mirror — the same slope-and-wedge reasoning read from the buy side of the labour market.

Structural Tensions

T1: Power versus size (the slope, not the share). Market power's defining move severs a large firm from a powerful one: what matters is the slope of the residual demand curve, not market share. A dominant-share firm hemmed in by close substitutes can raise price and lose nearly all its sales (little power), while a small differentiated firm with a captive base can raise price and keep most of them (real power). This is the concept's sharpest and most counterintuitive edge — and its practical hazard runs both ways. Reading share as power over-flags large benign firms; ignoring share entirely can miss that concentration is often how the slope gets steep. The size-independent test is liberating precisely because it refuses the share heuristic, which is also what makes it harder to apply than counting market shares. Diagnostic: Can this firm profitably raise price above cost without losing all custom (power), regardless of whether its share is large or small?

T2: Markup as diagnostic versus markup from other sources (a positive wedge need not be power). The Lerner index (P − MC)/P is the canonical meter of market power, positive wherever power exists. But a positive price-cost wedge is not identical to the ability to profitably restrict output: it can reflect recovery of large fixed costs, genuine scarcity, or risk premia, so the diagnostic can read "power" where the structural condition — an exploited downward-sloping residual demand — is absent. The problem deepens because marginal cost is notoriously unobservable, so measured markups rest on proxies, and the macro claim that aggregate markups rose after 1980 inherits exactly this fragility. The scalar that makes power quantifiable and comparable is only as trustworthy as the marginal-cost estimate beneath it and the assumption that the wedge reflects restriction rather than cost recovery. Diagnostic: Does this markup reflect an exploited downward-sloping residual demand (power), or fixed-cost recovery, scarcity, or a mismeasured marginal cost?

T3: Structural finding versus verdict of wrongdoing (having power is not abusing it). Possessing market power is a structural fact — a firm faces a sloped residual demand — not by itself an antitrust violation; firms acquire it through superior products, innovation, or scale, all lawful. The welfare concern attaches to its exercise and to the conduct that entrenches it, so "has market power" and "is behaving anticompetitively" are distinct claims. The tension is that the same measurement that establishes the structural finding is often marshalled as if it settled the normative one, and enforcement must hold the line: a high Lerner index is evidence of a condition, not a finding of harm. Collapsing the two either condemns successful firms for their success or excuses genuine abuse because power was "earned." Diagnostic: Is the claim that the firm possesses market power (a structural finding) or that it exercises or entrenches it harmfully (the welfare or legal verdict)?

T4: Distributional versus efficiency effect (one wedge, two consequences, different remedies). Any positive price-cost wedge licenses two simultaneous but distinct inferences: a distributional transfer of surplus from buyers (or, in monopsony, workers) to the powerful side, and an efficiency loss — mutually profitable trades that do not happen, recorded as the deadweight triangle. The concept's discipline is to keep them apart, because the policy responses differ: a transfer might be met with redistribution or price regulation, a deadweight loss with measures that restore the missing trades. The tension is that a single markup is both at once, so treating it as merely a transfer to redistribute ignores the trades destroyed, while treating it as merely lost efficiency ignores who gained the surplus — and a remedy aimed at one consequence need not address the other. Diagnostic: Is the concern the surplus transferred to the powerful side (distributional) or the trades never made at the above-cost price (efficiency) — and does the proposed remedy target the right one?

T5: Autonomy versus reduction (a price-theoretic concept or the bargaining-power/bottleneck parents). Within economics market power transfers as mechanism: product-market seller power and labour-market buyer power (monopsony) are genuine co-instances of one sloped-response-curve structure, handled with a single toolkit — residual demand, marginal cost, the Lerner wedge, deadweight loss — from industrial organization to antitrust to the macro markup literature. But cross-domain "platform power" or "narrative power" rest on different structural commitments; invoking "market power" for them imports specifically economic apparatus (the demand curve, marginal cost, the welfare triangle) that does not translate without distortion. What genuinely travels is the more general pattern — bargaining_power (one side's outside option is worse) and bottleneck (control of an essential input). The tension is between a price-theoretic specialization whose apparatus earns its own study and the recognition that its cross-substrate lesson belongs to those parents. Diagnostic: Resolve toward bargaining_power/bottleneck when reaching beyond priced markets; toward market power when the residual demand curve, marginal cost, and deadweight-loss accounting are literally in play.

Structural–Framed Character

Market power sits at the framed-leaning position on the structural–framed spectrum, held off the framed pole by a genuine structural core and a deliberately non-normative stance, but pushed onto the framed side by being wholly constituted by market institutions. On evaluative_weight it is low, and unusually self-aware about it: the entry insists that "having market power is a structural finding, not a verdict of wrongdoing" — possessing power is a lawful condition, and the concept meters a capacity (the price-cost wedge) rather than convicting a move, explicitly holding the structural finding apart from the welfare or legal verdict. That deliberate neutrality is its strongest structural mark. But human_practice_bound is high: the whole object — a firm facing a downward-sloping residual demand curve, a price above marginal cost, surplus transferred, a deadweight-loss triangle — presupposes markets, prices, firms, and buyers, and dissolves the instant that economic practice is removed; there is no observer-free market power, no monopsony in nature. Institutional_origin is moderate: the capacity to set price above the competitive level is a real feature of actual markets, but the apparatus that defines and measures it (the Lerner index, residual-demand estimation, the deadweight-loss accounting, the Cournot/Bertrand oligopoly models) is price-theoretic furniture of a specific economic tradition. Vocab_travels is low: residual demand, marginal cost, the Lerner wedge, elasticity, and deadweight loss are economics terms that lose their referents off priced markets. On import_vs_recognize the pattern is bimodal but tips framed at the boundary that matters: within economics, product-market seller power and labour-market monopsony are genuine co-instances of one sloped-response-curve structure recognized with a single toolkit, but beyond it — "platform power," "narrative power" — the look-alikes rest on different structural commitments and travel under the parents, so importing the market-power machinery there distorts.

The portable structural skeleton is the general sloped-response-curve surplus extraction: an actor faces a sloped rather than flat response curve to its own choices because the other side's alternatives are imperfect substitutes, so it can extract surplus by moving along the curve — carried at the prime level by bargaining_power (one side's outside option is worse) and bottleneck (control of an essential input). That skeleton is substrate-portable and is what genuinely recurs across domains. But it does not pull market power off the framed side, because that portable structure is precisely what market power instantiates from those parents as its price-theoretic specialization, not what makes "market power" itself travel: the cross-substrate reach belongs to bargaining-power and bottleneck control, while the residual demand curve, marginal cost, the Lerner index, the monopsony mirror, and the deadweight-loss triangle are exactly the domain apparatus that stays home in priced markets. Its character: an evaluatively neutral, deliberately non-normative economic property that is nonetheless wholly market-institution-constituted, structural only in the sloped-response-curve surplus-extraction skeleton it borrows from bargaining_power/bottleneck and specializes with demand-curve-and-marginal-cost machinery.

Structural Core vs. Domain Accent

This section settles why market power is a domain-specific abstraction rather than a prime, and it carries the case for its domain-specificity — so it is worth being exact about what could lift and what stays in priced markets.

What is skeletal (could lift toward a cross-domain prime). Strip the price theory and a thin relational structure survives: an actor faces a sloped — not flat — response curve to its own choices because the other side's alternatives are imperfect substitutes, so it can extract surplus by moving along the curve. The pieces that travel are abstract — an actor whose choices meet a finite rather than infinitely elastic response, a substitution friction that produces the slope by denying the other side a costless exit, and the surplus extraction that the slope permits. This skeleton is genuinely substrate-portable, which is exactly why it is carried at the prime level by bargaining_power (one side's outside option is worse than the other's) and bottleneck (control of an essential input conferring an analogous extractive position), and its recurrence is mechanism, not metaphor. But it is the core market power shares, not what makes it distinctive.

What is domain-bound. Almost everything that makes the concept market power in particular is price-theoretic furniture, and none of it survives extraction. The downward-sloping firm-level residual demand curve (and its upward-sloping labour-supply mirror); marginal cost and the price–marginal-cost wedge metered by the Lerner index (P − MC)/P with its reciprocal-of-elasticity dual; the distributional and efficiency welfare readouts, the latter recorded as the deadweight-loss triangle; the monopsony mirror read from the buy side as wages below marginal revenue product; the aggregation property carrying the scalar up to economy-wide markups, labour share, and productivity dispersion; and the operational apparatus (HHI, SSNIP/diversion ratios, the Cournot/Bertrand/Stackelberg models, antitrust and rate regulation). These are the worked vocabulary, the instruments, and the empirical literature specific to priced markets. The decisive test: remove prices, firms, marginal cost, and surplus accounting, and the residual demand curve, the Lerner wedge, and the deadweight-loss triangle have nothing to refer to; what remains is a bare sloped-response-curve extraction, a looser thing that is bargaining_power/bottleneck, not market power.

Why this does not clear the prime bar. A prime is a relational structure whose vocabulary travels and whose cross-domain transfer is recognition of the same mechanism, not analogy. Market power's transfer is bimodal. Within economics it travels intact — the size-independent test, the Lerner quantification, the source-decomposition of the slope, and the two-branch welfare readout move without translation across industrial organization, antitrust, labour economics (the monopsony mirror), regulation, and the macro markup literature, because product, labour, capital, and platform markets are co-instances of one sloped-response-curve structure. Beyond it — "platform power," "narrative power" — it travels only by renaming the residual curve, the friction, and the surplus while importing the demand-curve-and-marginal-cost machinery that does not translate; these look-alikes rest on different structural commitments and are analogies, not the same mechanism. And when the bare structural lesson is genuinely needed cross-domain — an actor extracts surplus because the other side lacks close substitutes — it is already carried, in more general form, by the parent primes market power instantiates: bargaining_power and bottleneck. The cross-domain reach belongs to those parents; "market power," as named, carries residual-demand-and-deadweight-loss baggage that does not and should not travel.

Relationships to Other Abstractions

Local relationship map for Market powerParents appear above the current abstraction, mutual partners to the right, and children below. Node labels state whether each abstraction is prime or domain-specific; colors identify relation types.Market powerDOMAINPrime abstraction: Bargaining Power — is a decomposition ofBargaining PowerPRIMEPrime abstraction: Positional Advantage — is a decomposition ofPositionalAdvantagePRIMEDomain-specific abstraction: Monopolistic Competition — is part ofMonopolisticCompetitionDOMAINDomain-specific abstraction: Oligopoly — is part of, conditionalOligopolyDOMAINDomain-specific abstraction: Lerner index — presupposesLerner indexDOMAINDomain-specific abstraction: Monopsony power — is a kind ofMonopsony powerDOMAIN

Current abstraction Market power Domain-specific

Parents (2) — more general patterns this builds on

  • Market power is a decomposition of Bargaining Power Prime

    Market power is the price-theoretic specialization of bargaining power: finite substitution makes the counterparty's exit option costly and lets the actor move terms and surplus toward itself.

  • Market power is a decomposition of Positional Advantage Prime

    Removing price theory leaves an actor whose occupied position among alternatives confers terms-setting leverage independent of its absolute size.

Children (4) — more specific cases that build on this

  • Monopsony power Domain-specific is a kind of Market power

    Monopsony power is market power specialized to a buyer facing an upward-sloping residual supply curve and extracting a markdown.

  • Lerner index Domain-specific presupposes Market power

    The Lerner Index presupposes market power as the target attribute whose exercised price-cost wedge it maps onto a scale.

  • Monopolistic Competition Domain-specific is part of Market power

    Every firm in monopolistic competition contains local market power over its differentiated variety, expressed by downward-sloping residual demand and price above marginal cost.

Hierarchy paths (2) — routes to 1 parentless root

Not to Be Confused With

  • Market share / concentration (bigness). How large a firm is, or how concentrated its market (e.g. HHI). Market power is read off the slope of the residual demand curve, not the share: a dominant-share firm hemmed in by close substitutes has little power (raise price, lose nearly all sales), while a small differentiated firm can have real power. Concentration is often how the slope gets steep, but it is not the power itself. Tell: is the claim about how big the firm's share is (concentration), or whether it can profitably raise price above cost without losing all custom (market power)?

  • Accounting markup (price above marginal cost, taken alone). A positive P − MC is the diagnostic symptom, but a markup can reflect recovery of large fixed costs, genuine scarcity, or risk premia rather than the ability to profitably restrict output. The structural condition is an exploited downward-sloping residual demand, not any price-cost gap in isolation. Tell: does the markup come from exploiting a sloped residual demand (power), or from fixed-cost recovery, scarcity, or a mismeasured marginal cost (not necessarily power)?

  • The Lerner index. Market power's own meter — (P − MC)/P, equal to the reciprocal of residual-demand elasticity — not the thing itself. The index is the readout; the power is the underlying capacity to set price above the competitive level. A measured markup is evidence of power, not identical to it (and rests on an unobservable marginal cost). Tell: is the object the scalar that quantifies the capacity (Lerner index), or the capacity it quantifies (market power)?

  • Monopoly. A market structure — a single seller. Market power is a graded capacity to price above the competitive level that exists in degrees wherever residual demand slopes down, possessed by differentiated competitors and oligopolists, not only monopolists; and a monopolist facing very elastic substitutes can have little of it. Tell: is the claim about how many sellers there are (monopoly = one), or about the degree of the price-cost wedge a firm can sustain (market power, a continuum)?

  • Bargaining power / bottleneck (the parents — and "platform/narrative power"). The substrate-neutral patterns market power specializes — one side extracting surplus because the other's outside option is worse (bargaining_power), or control of an essential input (bottleneck). Cross-substrate look-alikes ("platform power," "narrative power") rest on different structural commitments and travel under these parents, not under the residual-demand-and-deadweight-loss apparatus; importing the economic machinery there distorts. Tell: strip the residual demand curve, marginal cost, and welfare triangle and what remains — surplus extraction from imperfect substitutes — is bargaining_power/bottleneck (treated more fully in Structural Core vs. Domain Accent); "market power" is present only where priced markets with marginal cost are literally in play.

Neighborhood in Abstraction Space

Market power sits in a crowded region of the domain-specific corpus (11th percentile for distinctiveness): several abstractions share nearly its structure, so a description that fits it tends to fit its neighbors too.

Family — Market Structure & Price Equilibrium (25 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-07-12