Monopsony power¶

Gauge a buyer's ability to set the price it pays below the competitive level by the slope of the upward-sloping residual supply curve it faces (finite elasticity ε), which yields a markdown of roughly 1/ε and the double distortion of underpayment plus under-hiring.

Core Idea¶

Monopsony power is the buyer-side analog of monopoly: a buyer's ability to set the price it pays below the competitive level and sustain that markdown without losing all its suppliers. The structural signature is an upward-sloping residual supply curve faced by the buyer — to attract one more seller (most importantly, one more worker), the buyer must raise the price it pays to every seller already in the pool. The marginal cost of acquiring an additional unit therefore exceeds the price actually paid, and profit-maximizing behavior equates the buyer's marginal revenue product of that unit to the higher marginal cost rather than to the price, leaving sellers with compensation strictly below their marginal product.

The pure-monopsony case — one buyer, many sellers — is rare. The empirically operative concept is monopsony power on a spectrum: any friction that makes an individual seller's supply elasticity to a particular buyer finite generates some degree of buyer-side pricing power. Commuting costs, search frictions, occupational licensing, non-compete clauses, tied immigration status, and employer concentration all limit how readily a worker can move to an alternative employer; each source of friction steepens the residual supply curve facing the incumbent buyer and widens the markdown. Alan Manning's 2003 synthesis Monopsony in Motion showed that ordinary search frictions are sufficient to generate substantial markdowns even in apparently competitive labor markets, making monopsony power a standard condition rather than a rare exception.

The welfare consequence is a double distortion: workers are paid less than their marginal product, and employment is held below the competitive level because the monopsonist restricts hiring to avoid raising wages to the full pool. This creates the counterintuitive policy result that minimum wage floors — up to some level — can raise both wages and employment simultaneously in monopsonistic labor markets, since the floor prevents the buyer from exploiting the inframarginal markup on existing workers (the Card–Krueger empirical result in fast food is standardly interpreted through this lens). Interventions that raise workers' outside options — non-compete bans, no-poach prohibitions, portability of employer-sponsored benefits, competing employers in the region — flatten the residual supply curve and reduce the markdown.

Structural Signature¶

Sig role-phrases:

the buyer with power — an employer (or processor, platform) able to set the price it pays below the competitive level without losing all its suppliers
the suppliers — the workers/sellers whose outside option to this particular buyer is weak, so they cannot all walk away
the upward-sloping residual supply curve — the structural signature: attracting one more seller requires raising the price paid to every seller already in the pool, summarized by a finite firm-level supply elasticity ε
the friction sources — the additive contributors to that slope: commuting cost, search friction, occupational licensing, non-compete clauses, tied immigration status, employer concentration
the marginal-cost-exceeds-price wedge — because the raise applies to all, the marginal cost of one more unit exceeds the price actually paid
the markdown — profit-maximization equates marginal revenue product to that higher marginal cost, leaving compensation below marginal product by roughly 1/ε
the double distortion — pay below marginal product and hiring held below the competitive level at once, so flattening the curve relaxes both together

What It Is Not¶

Not confined to the single-buyer ("one company town") case. Pure monopsony is rare; the operative concept is power on a spectrum. Any friction that makes a worker's supply elasticity to a particular employer finite — commuting, search costs, licensing, non-competes, tied visas, concentration — confers some markdown. Manning's result is precisely that ordinary search frictions suffice to generate substantial monopsony power in apparently competitive markets, so the absence of a literal sole employer is no defense.
Not established by low wages alone. A market can pay little because marginal product is genuinely low (competitive wage, infinite elasticity, no markdown) or because the buyer extracts a wedge below marginal product (finite elasticity, positive markdown). The two cases can post identical wages; what diagnoses monopsony is the slope of the residual supply curve, not the level of pay. Reading every low wage as evidence of buyer power skips the structural test that the concept exists to impose.
Not a moral charge of exploitation. The concept converts "workers are underpaid because exploited" — which gives a labor economist no handle — into an estimable property: the firm faces supply elasticity ε, implying a quantifiable markdown between marginal product and wage. It locates and measures a distortion; it does not pronounce on fairness, and treating the term as a synonym for "the boss is greedy" discards exactly the measurement that makes it useful.
Not a guarantee that a minimum wage always raises employment. The counterintuitive Card–Krueger result is regime-conditional, not a universal law. Only where the buyer was restricting hiring to hold down the inframarginal wage (finite ε) can a floor that caps that markup raise wages and employment together; where the market is effectively competitive (large ε), a binding floor still moves employment down the demand curve. The sign of the effect depends on which side of the elasticity threshold the market sits.
Not oligopsony, and not employer collusion. Monopsony power is a property of the supply curve an individual buyer faces — it can arise from commuting or search frictions with many independent employers present. Oligopsony is the few-buyer case, and no-poach or wage-fixing agreements are one source of buyer power (or its own antitrust offense), not the phenomenon itself. The markdown can be large with no collusion at all.

Scope of Application¶

Monopsony power lives across the subfields of economics where a buyer faces an upward-sloping residual supply curve — a finite supply elasticity to that particular buyer; its reach is within that domain, with looser "platform/data monopsony" extensions belonging to the parent bargaining_power rather than to the elasticity-and-markdown apparatus itself.

Labor economics — the home turf, where search frictions, commuting, occupational licensing, non-competes, tied visas, and employer concentration are the catalog of supply-curve slope sources behind wage stagnation, large employer wage premia, and the minimum-wage puzzle.
Antitrust enforcement — no-poach and wage-fixing agreements treated as buyer-side offenses, and merger review extended to labor-market effects, applying product-market monopoly analysis to the buyer side.
Healthcare procurement — a rural hospital facing local nurses, or an insurer facing physicians, as a monopsonist in the model's exact sense, with the same elasticity and markdown driving provider pay.
Agricultural procurement — a regional meatpacker or dairy processor facing producers (the structure the 1921 Packers and Stockyards Act addresses), the same markdown with "farmers" in the seller role.
Sports and entertainment labor — reserve clauses, drafts, and salary caps as formal, engineered monopsony, historically the canonical cases prompting antitrust exemptions and collective-bargaining responses.
Migrant and unfree labor — tied-visa programs (H-2A, H-2B, kafala) that create extreme buyer power by collapsing the worker's outside option of switching employers to near zero.

Clarity¶

Naming monopsony power converts "why are wages low?" from a moral or political accusation into a measurement question. "Workers are underpaid because they are exploited" has no handle a labor economist can grip; "the firm faces a residual labor-supply elasticity of ε, implying a markdown of roughly 1/ε between marginal product and wage" is estimable, falsifiable, and tells you exactly how much pay is at stake. The concept localizes the cause of low pay to a single, observable property — the slope of the supply curve the individual employer faces — and so separates a market that merely pays little (low marginal product, competitive wage) from one that pays below marginal product (positive markdown, finite elasticity). That distinction is invisible without the concept and decisive for policy.

It also dissolves the textbook reflex that a wage floor must cost jobs. Under perfect competition a binding minimum wage moves the market up a downward-sloping demand curve and employment falls; the monopsony diagnosis shows why this need not hold — when the buyer was restricting hiring precisely to avoid bidding up the wage on its inframarginal workers, a floor that caps that markup can raise wages and employment together. The sharper question the practitioner can now ask is not "is intervention good or bad for workers?" but "which source of supply-curve slope does this intervention target — commuting, search, licensing, non-competes, tied visas, employer concentration — and how much elasticity does flattening it buy?" Each policy debate (minimum wage, no-poach bans, benefit portability, merger review for labor effects) becomes a question about a specific friction and its elasticity-shift, rather than a clash of intuitions about employer fairness.

Manages Complexity¶

Low and stagnant wages, the puzzle of minimum-wage floors that fail to destroy jobs, no-poach suits, the wage effects of mergers, the bite of non-compete clauses, occupational licensing, tied-visa exploitation, gig-platform reliance — each arrives in the labor literature as its own debate with its own facts and its own contested remedy. Monopsony power compresses that sprawl to a single property of one market: the slope of the residual supply curve the individual buyer faces, summarized by one number, the firm-level supply elasticity ε. From ε the rest is read off mechanically. The markdown — the wedge by which pay falls below marginal product — is roughly 1/ε; the direction and rough size of every wage shortfall follow from that one estimate rather than from a separate story about employer fairness in each case. The analyst no longer asks "is this market exploitative?" good or bad, but locates the cause of low pay in a single observable quantity, and that quantity at once separates a market that merely pays little because marginal product is low (competitive wage, ε infinite, no markdown) from one that pays below marginal product (finite ε, positive markdown) — a distinction invisible without the concept and decisive for whether intervention can help at all.

The many policy debates then collapse into one structural question with a uniform shape: which friction is steepening this buyer's supply curve — commuting cost, search friction, occupational licensing, non-compete, tied immigration status, employer concentration — and how much does flattening it lower ε? Each named friction is an additive contributor to the slope, so the catalog of interventions (minimum-wage floors, no-poach bans, benefit portability, merger review for labor effects, more competing employers) sorts by which friction it targets and how much elasticity the flattening buys, rather than by clashing intuitions. The branch structure is the elasticity itself: where ε is large the market is effectively competitive and a wage floor moves employment down the old demand curve and destroys jobs; where ε is finite the buyer was already restricting hiring to hold down the inframarginal wage, so a floor capping that markup raises wages and employment together — the entire counterintuitive minimum-wage result drops out of which region of ε the market sits in. A high-dimensional catalog of "which labor-market complaints are real and which fixes work" reduces to estimating one slope and reading the welfare and policy consequences off its value.

Abstract Reasoning¶

Monopsony power licenses reasoning that locates the cause of low pay in one observable property — the slope of the residual supply curve an individual buyer faces — and reads diagnosis, welfare, and policy off the firm-level supply elasticity ε.

The foundational move is converting an accusation into a measurement. Confronting "why are wages low?", the analyst refuses the moral framing ("workers are exploited," which has no handle) and reasons to an estimable property: the firm faces a residual labor-supply elasticity of ε, implying a markdown of roughly 1/ε between marginal product and wage. The reasoning runs from a structural slope to a quantified pay shortfall, so the question becomes falsifiable and the dollars at stake become computable. This is the move that makes "underpaid" a claim a labor economist can grip rather than assert.

The decisive diagnostic move is separating "pays little" from "pays below marginal product". The analyst uses ε to distinguish two markets that look alike on the wage but differ structurally: where ε is effectively infinite the market is competitive and a low wage merely reflects low marginal product (no markdown, nothing to fix); where ε is finite the buyer pays below marginal product (positive markdown, surplus extracted). The reasoning infers from the supply-curve slope which case obtains, and that distinction is decisive for whether intervention can help at all — invisible without the concept, because the two cases can post identical wages.

A third move is interventionist friction-targeting. The analyst reasons that ε is not primitive but the sum of named frictions — commuting cost, search friction, occupational licensing, non-compete clauses, tied immigration status, employer concentration — each an additive contributor to the supply curve's slope. So every policy debate sorts by which friction it attacks and how much flattening it buys: non-compete bans, no-poach prohibitions, benefit portability, merger review for labor effects, more competing employers in the region. The reasoning runs from a target friction to a predicted reduction in ε to a predicted shrinkage of the markdown, replacing clashing intuitions about employer fairness with a question about a specific friction and its elasticity-shift.

The fourth and most counterintuitive move is regime-conditional prediction of a minimum wage's effect. The analyst reasons that the employment consequence of a wage floor depends on which region of ε the market sits in. Where ε is large (competitive), a binding floor moves the market up a downward-sloping demand curve and employment falls — the textbook reflex. Where ε is finite (monopsonistic), the buyer was already restricting hiring precisely to avoid bidding up the inframarginal wage, so a floor that caps that markup can raise wages and employment together. The entire Card–Krueger result drops out not as an anomaly but as the prediction for the finite-ε region, and the reasoning is to establish the regime before predicting the sign — the same intervention has opposite employment effects on the two sides of the elasticity threshold.

Underwriting all of these is the marginal-cost-exceeds-price move that generates the structure. Because attracting one more seller requires raising the price paid to every seller already in the pool, the analyst reasons that the marginal cost of an additional unit exceeds the price actually paid, so profit-maximization equates marginal revenue product to that higher marginal cost rather than to the price — leaving compensation strictly below marginal product and employment below the competitive level. This is the inference that yields the double distortion (underpayment and under-hiring at once), and it is what lets the analyst predict that flattening the supply curve relaxes both distortions together rather than trading one off against the other.

Finally, the concept supports a mirror-and-transfer move. The analyst reasons that monopsony is the buyer-side reflection of monopoly — markdown for the buyer where the seller has a markup — so the symmetric apparatus carries to any procurement setting with an upward-sloping residual supply curve: a meatpacker facing regional cattle producers, a hospital facing local nurses, a platform facing single-platform gig workers, a tied-visa employer whose worker's outside option collapses to zero. The move is to recognize finite supply elasticity to a particular buyer wherever it arises and apply the same slope-and-markdown reasoning, while noting that analogical extensions (platform, data monopsony) lack the labor-supply-elasticity apparatus that gives the diagnosis its quantitative force.

Knowledge Transfer¶

Within economics the concept transfers as mechanism, and its precondition is precise and portable: anywhere a buyer faces an upward-sloping residual supply curve — finite supply elasticity to that particular buyer — the whole apparatus moves intact. The diagnostic (estimate ε), the welfare reading (markdown ≈ 1/ε, the double distortion of underpayment plus under-hiring), the intervention logic (find the friction steepening the curve and flatten it), and the regime-conditional minimum-wage prediction all carry across subfields without translation, because each subfield is a genuine instance of the same buyer-side pricing structure, not a likeness of it.

Across the home domain this is a literal, mechanism-preserving transfer. In labor economics it is the home turf — search frictions, commuting, licensing, non-competes, tied visas, and employer concentration are the catalog of slope sources. In healthcare procurement a rural hospital facing local nurses, or an insurer facing physicians, is a monopsonist in exactly the model's sense, with the same elasticity and markdown. In agricultural procurement a regional meatpacker facing cattle producers (the structure the 1921 Packers and Stockyards Act addresses) is the same buyer-side markdown with "workers" replaced by "farmers" but the apparatus unchanged. In sports and entertainment reserve clauses, drafts, and salary caps are formal, engineered monopsony — historically the canonical cases. And antitrust enforcement of no-poach and wage-fixing agreements, and labor-effects merger review, is itself a within-discipline transfer of product-market monopoly analysis to the buyer side. In each, "buyer," "seller," and "outside option" are re-instantiated but the marginal-cost-exceeds-price machinery is the same machinery.

Beyond that range the honest report splits in two. (1) The named apparatus — residual labor-supply elasticity, the marginal cost of labor, the MRP-minus-wage wedge, the 1/ε markdown — does not travel; it is bound to settings with a measurable supply curve to an individual buyer. Invocations like "platform monopsony" over content creators or "data monopsony" over user attention have genuine structural force but are analogy at the apparatus level: they borrow the buyer-has-power shape while lacking the labor-supply-elasticity and marginal-cost-of-input machinery that gives the diagnosis its quantitative bite, and the honest move is to mark them so. (2) What genuinely recurs across domains is one level up: the general pattern of imperfect competition / asymmetric bargaining power — one party can move price against another because the other's outside options are weak. That parent really does travel, and travels as mechanism, to product, labor, and even political "markets." Monopsony is its buyer-side, supply-curve-equipped specialization. So when the cross-domain lesson is needed, it should carry the parent — bargaining_power / the buyer-side mirror of market_power — not the name "monopsony," whose elasticity apparatus, markdown formula, and friction catalog are labor-economics-and-procurement furniture that does not and should not travel. This is exactly the mechanism-within / analogy-(or parent-)beyond boundary that Structural Core vs. Domain Accent develops.

Examples¶

Canonical¶

When New Jersey raised its state minimum wage from $4.25 to $5.05 an hour on April 1, 1992, David Card and Alan Krueger surveyed 410 fast-food restaurants (Burger King, KFC, Wendy's, Roy Rogers) in New Jersey and neighboring eastern Pennsylvania, before and after the change. The competitive textbook predicted employment in New Jersey should fall relative to the Pennsylvania control. Instead, New Jersey employment did not decline — if anything it rose slightly relative to Pennsylvania. Read through the monopsony lens, this is exactly the prediction for the finite-elasticity regime: fast-food employers had been holding hiring below the competitive level to avoid bidding up wages on their existing crews, so a floor capping that inframarginal markup raised pay while leaving employment intact — the anomaly that put buyer power in low-wage labor markets on the empirical map.

Mapped back: The restaurants are the buyer with power, and their reluctance to raise crew wages reveals the upward-sloping residual supply curve (a finite ε) they face. Holding staffing below the competitive level to protect the inframarginal wage is the markdown and the double distortion operating together. The floor's employment-neutral effect is the regime-conditional signature: the same intervention that destroys jobs where ε is large leaves them intact where ε is finite.

Applied / In Practice¶

Douglas Staiger, Joanne Spetz, and Ciaran Phibbs exploited a natural experiment in the U.S. Department of Veterans Affairs hospital system, where a statutory change to how VA nurse wages were set shifted pay at some hospitals, to estimate the labor-supply elasticity that a single hospital actually faces. If nurses moved freely to competing employers, a hospital changing its wage would see employment swing sharply — elastic supply. Instead they found the short-run supply of nurses to an individual hospital to be strikingly inelastic, an elasticity on the order of 0.1, implying substantial buyer-side markdown power even in cities with many hospitals. The result reframes chronic nurse "shortages" as a symptom of monopsony: hospitals rationally hold both nurse wages and staffing below competitive levels.

Mapped back: Nurses are the suppliers whose outside option to a given hospital is weak. The directly estimated low elasticity is the upward-sloping residual supply curve made quantitative, and commuting and thin local hospital markets are the friction sources steepening it. A supply elasticity near 0.1 implies a large markdown (≈ 1/ε), and the persistently below-competitive staffing is the double distortion.

Structural Tensions¶

T1: Measurement versus moral charge (the analytic handle bought by dropping fairness). The concept's signal move is converting "workers are exploited" into "the firm faces elasticity ε, so the markdown is roughly 1/ε" — trading a claim with no economist's handle for one that is estimable and falsifiable. The gain is real: dollars at stake become computable and policy debates sort by friction. But the same move strips out the normative content that motivated the question. A distributional or political grievance can be laundered into a technical elasticity estimate that settles the size of the wedge while remaining silent on whether any wedge is legitimate; conversely, workers experiencing underpayment as injustice can find the diagnosis dismissive precisely because it "only measures." The concept sharpens what it can quantify by refusing what it cannot. Diagnostic: Is the dispute here about the measurable magnitude of the markdown, or about whether a markdown of any size is acceptable — a question ε does not answer?

T2: Below marginal product versus merely low pay (the decisive distinction the surface wage hides). The entire diagnostic rests on separating a market that pays little because marginal product is low (competitive, ε infinite, nothing to fix) from one that pays below marginal product (finite ε, positive markdown). This distinction is decisive for whether intervention can help at all — and it is invisible on the wage itself, since the two cases can post identical pay. The whole weight of the concept therefore falls on the slope of the residual supply curve, the one thing hardest to observe directly: credibly estimating ε requires exogenous wage variation (the NJ border, the VA statutory change), which clean natural experiments alone supply. The concept's central claim is simultaneously its most decisive and its least directly readable feature. Diagnostic: Is ε here identified off exogenous variation, or is "monopsony" inferred from the low wage alone — the very move the concept forbids?

T3: Ubiquity versus discriminating power (Manning's spectrum). Manning's synthesis showed that ordinary search frictions suffice to generate substantial markdowns even in apparently competitive markets, making monopsony power "a standard condition rather than a rare exception." That universality is the empirical triumph of the concept — it dissolves the "one company town" caricature. But it also erodes the diagnosis's sharpness: if every buyer faces some finite ε, then "there is monopsony power here" is nearly always true and stops distinguishing one market from another. The concept survives only by shifting from a yes/no verdict to a magnitude question — how large is the markdown — where the near-universal presence of some friction carries no information on its own. Its reach and its bite pull against each other. Diagnostic: Does the claim specify how much markdown is at stake, or does it rest on the near-tautology that some friction exists?

T4: Regime-conditional, non-monotone minimum wage (the showpiece prediction that flips sign). The counterintuitive Card–Krueger result — a floor raising wages and employment together — is the concept's most celebrated payoff, but it is knife-edged twice over. The employment effect flips sign across the elasticity threshold: where ε is large the same floor moves employment down the demand curve and destroys jobs. And even within the monopsony region the benefit is non-monotone — a floor helps only up to some level, beyond which it recreates the competitive job-loss region. A policymaker who misreads the regime, or sets the floor too high, gets exactly the textbook harm the result seemed to refute. The boldest prediction the concept licenses is also the one most punished by getting the parameters wrong. Diagnostic: Have the regime (finite versus large ε) and the ceiling below which the floor helps both been established, or is "minimum wages raise employment" being applied unconditionally?

T5: No equity–efficiency tradeoff versus the standard regime (the double distortion's joint relaxation). Because underpayment and under-hiring flow from a single wedge, an intervention that flattens the supply curve improves both margins at once — dissolving the canonical equity-efficiency tradeoff in which raising pay by fiat costs jobs. This joint relaxation is genuinely unusual and is much of the concept's welfare appeal. But it is not the abolition of the tradeoff; it is the identification of the one regime where the tradeoff is suspended. Outside the finite-ε region the ordinary tradeoff snaps back. The danger is that the attractive "win-win" framing travels ahead of the regime evidence, importing a free-lunch promise into markets whose large ε will punish it. Diagnostic: Is the claimed win-win backed by evidence the market sits in the finite-ε region, or is a monopsony free lunch being asserted where the standard tradeoff may still rule?

T6: Autonomy versus reduction (its own quantified apparatus or an instance of bargaining power). "Monopsony power" is a fully furnished, canonically studied construct — residual supply elasticity, marginal cost of labor, the 1/ε markdown, the friction catalog — and that apparatus is exactly what gives the diagnosis its quantitative bite. Yet the apparatus is bound to settings with a measurable supply curve to an individual buyer; it does not travel. What recurs across product, labor, and even political "markets" is one level up: imperfect competition / asymmetric bargaining power, the general fact that a party can move price when the other's outside options are weak. "Platform monopsony" and "data monopsony" borrow the buyer-has-power shape while lacking the elasticity machinery, so they are the parent wearing the specialized name. The tension is between a labor-economics construct that earns its own quantitative study and the recognition that its cross-domain cargo already belongs to bargaining_power. Diagnostic: Resolve toward the parent (bargaining power, the buyer-side mirror of market power) when asking what travels outside settings with a measurable supply curve; toward named monopsony when quantifying a specific buyer's markdown in situ.

Structural–Framed Character¶

Monopsony power is mixed on the structural–framed spectrum — an evaluatively neutral (indeed deliberately de-moralized) quantified market property with a clean portable bargaining-power skeleton, but one constituted by market institutions and stated in labor-economics vocabulary, so it holds the middle. The criteria: on evaluative weight it is pointedly structural, and the entry makes this a load-bearing feature — the concept's signature move is to convert "workers are underpaid because exploited" (a moral charge with no analytic handle) into "the firm faces supply elasticity ε, implying a markdown of roughly 1/ε," an estimable property; it locates and measures a distortion without pronouncing on fairness, explicitly declining the exploitation verdict. But human-practice-bound points framed: monopsony power exists only inside a market with buyers, sellers, wages, and outside options — human-institutional throughout; there is no monopsony in observer-free nature. Institutional origin is mixed: the residual-supply-curve slope is a real structural property, but it is a property of human labor and procurement markets, and its friction sources (occupational licensing, non-competes, tied-visa programs) are themselves institutional artifacts. Vocab-travels is low: residual supply elasticity, marginal cost of labor, the MRP-minus-wage wedge, the 1/ε markdown are labor-economics and procurement idiom. Import-vs-recognize is bimodal: within economics (labor, healthcare and agricultural procurement, sports labor, antitrust) it transfers as mechanism wherever a buyer faces a measurable upward-sloping supply curve; beyond that ("platform monopsony," "data monopsony") it is analogy at the apparatus level, borrowing the buyer-has-power shape without the elasticity machinery.

The portable structural skeleton is a single one: asymmetric bargaining power under imperfect competition — a party can move price against another because the other's outside options are weak. That skeleton genuinely travels to product, labor, and even political "markets," but it is exactly what monopsony power instantiates from its umbrella prime — bargaining_power (the buyer-side mirror of market_power) — not what makes "monopsony power" itself portable: the cross-domain reach belongs to that parent, of which monopsony is the buyer-side, supply-curve-equipped specialization, while the domain-accented cargo — the residual supply elasticity ε, the marginal-cost-exceeds-price wedge, the 1/ε markdown formula, the double distortion, the friction catalog, the regime-conditional minimum-wage result — stays home with settings that have a measurable supply curve to an individual buyer. Its character: a deliberately evaluatively-neutral, quantified market-power property, structural in the bargaining-power skeleton it borrows from market_power, but constituted by labor and procurement market institutions and stated in an elasticity-and-markdown vocabulary that pins the named concept to its home domain, leaving it mixed rather than a free-floating prime.

Structural Core vs. Domain Accent¶

This section settles why monopsony power is a domain-specific abstraction and not a prime.

What is skeletal (could lift toward a cross-domain prime). Strip the labor market away and a thin relational structure survives: asymmetric bargaining power under imperfect competition — one party can move the price against another because the other's outside options are weak, and that party sets terms in its own favor at the expense of the constrained side. The portable pieces are abstract — a party facing a counterpart whose ability to walk away is limited, the resulting capacity to push price off the competitive level, and the surplus extracted from the constrained side. That skeleton is genuinely substrate-portable, reaching product markets, labor, and even political "markets," which is exactly why the entry houses it in the umbrella prime monopsony instantiates: bargaining_power, the buyer-side mirror of market_power. But this is the core monopsony power shares, not what makes it monopsony power.

What is domain-bound. Everything that individuates the concept is labor-economics-and-procurement furniture that does not survive extraction: the residual supply elasticity ε as the one observable that summarizes the buyer's power; the upward-sloping residual supply curve whose slope is that power; the marginal-cost-exceeds-price wedge (raising the price to attract one more seller raises it for the whole inframarginal pool); the markdown formula (≈ 1/ε) converting the slope into a quantified pay shortfall; the double distortion of underpayment plus under-hiring flowing from a single wedge; the friction catalog (commuting cost, search friction, occupational licensing, non-competes, tied-visa status, employer concentration) that additively steepens the curve; and the regime-conditional minimum-wage result by which a floor raises wages and employment together only in the finite-ε region. The decisive test: remove a measurable supply curve to an individual buyer and none of this apparatus has anything to bind to — there is no elasticity to estimate, no markdown to compute, no regime to establish. What is left is the bare fact that one side has the whip hand, which is the parent, not the quantified child.

Why this does not clear the prime bar. A prime's vocabulary travels and its transfer is recognition of the same mechanism, not analogy; monopsony power's transfer is bimodal. Within economics it travels as literal mechanism wherever a buyer faces an upward-sloping residual supply curve — labor markets, a hospital facing local nurses, a meatpacker facing cattle producers, reserve-clause sports labor, antitrust's buyer-side offenses — the marginal-cost-exceeds-price machinery unchanged and only the identities of buyer, seller, and outside option re-instantiated. Beyond settings with a measurable supply curve it does not travel as a named unit: "platform monopsony" over content creators and "data monopsony" over user attention have real structural force but are analogy at the apparatus level, borrowing the buyer-has-power shape while lacking the supply-elasticity and marginal-cost-of-input machinery that gives the diagnosis its quantitative bite. When that cross-domain lesson is actually needed, it is already carried, in more general form, by bargaining_power (the buyer-side mirror of market_power), of which monopsony is the supply-curve-equipped specialization. The cross-domain reach belongs to that parent; the elasticity apparatus, the 1/ε markdown, the friction catalog, and the regime-conditional minimum-wage prediction are the domain-accented cargo that should stay home.

Relationships to Other Abstractions¶

Current abstraction Monopsony power Domain-specific

Parents (1) — more general patterns this builds on

Monopsony power is a kind of Market power Domain-specific

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

Hierarchy paths (2) — routes to 1 parentless root

Monopsony power → Market power → Bargaining Power → Asymmetry

Show alternative path (1)

Not to Be Confused With¶

Monopoly power (the sell-side mirror). A seller's ability to set price above the competitive level by facing a downward-sloping demand curve, earning a markup. Monopsony power is its exact buyer-side reflection — a buyer setting price below competitive by facing an upward-sloping supply curve, earning a markdown. The apparatus is symmetric (elasticity, wedge, distortion) but the side is opposite. Tell: does the power holder sell into a demand curve and mark price up (monopoly power), or buy from a supply curve and mark price down (monopsony power)?
Oligopsony. The few-buyer case, where a handful of buyers are large enough to interact strategically. Monopsony power is a property of the residual supply curve an individual buyer faces — it can arise from commuting or search frictions with many independent employers present, no strategic interaction required. Tell: does the pricing turn on strategic interaction among a few buyers (oligopsony), or on the finite supply elasticity to one buyer regardless of how many other buyers exist (monopsony power)?
Pure monopsony (the single-buyer "company town"). The literal one-buyer-many-sellers case. This is the rare limiting instance, not the operative concept: monopsony power is a spectrum generated by any friction that makes a seller's supply elasticity to a particular buyer finite (Manning's result), so the absence of a sole employer is no defense. Part-vs-whole: pure monopsony is the ε→0 corner of the broader power spectrum. Tell: is there literally one buyer (pure monopsony), or finite supply elasticity to a buyer amid many (the operative monopsony-power spectrum)?
Wage-fixing / no-poach collusion. Explicit agreements among employers not to compete for workers. These are one source of a steepened residual supply curve (and their own antitrust offense), not the phenomenon itself — the markdown can be large with no collusion at all, arising from commuting or search frictions among independent firms. Tell: is buyer power produced by an agreement among employers (collusion, a source), or by supply-curve frictions that would confer a markdown even absent any agreement (monopsony power proper)?
Exploitation (the moral charge, and Pigou's technical sense). The claim that "workers are underpaid because exploited." Monopsony power's signature move is to convert that unhandle-able moral charge into an estimable property — the firm faces elasticity ε, implying a markdown of roughly 1/ε — locating and measuring a distortion without pronouncing on fairness. Even where "exploitation" is used technically (Pigovian, pay below marginal product), monopsony supplies the elasticity apparatus that quantifies it. Tell: is the claim a normative verdict on fairness (exploitation), or a measured wedge between marginal product and wage read off the supply-curve slope (monopsony power)?
Bargaining power / market power (the umbrella). The substrate-general skeleton monopsony power instantiates — asymmetric bargaining power under imperfect competition, one party moving price because the other's outside options are weak. This parent (bargaining_power, the buyer-side mirror of market_power) is what carries the lesson beyond settings with a measurable supply curve; "platform monopsony" and "data monopsony" are analogy at the apparatus level, borrowing the buyer-has-power shape without the elasticity machinery. Tell: the umbrella (treated in a later section) is what travels; "monopsony power" as named applies only where a measurable residual supply curve to an individual buyer exists.

Neighborhood in Abstraction Space¶

Monopsony power sits in a crowded region of the domain-specific corpus (2^nd 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