RESEARCH DOSSIER

Production Costs and Supply Formation

Abstract. This research dossier examines the economic mechanisms by which production costs shape supply, participation, prices, and capacity adjustment. It integrates classical price theory, conditional input minimization, competitive output choice, imperfect competition, and empirical evidence on markups, pass-through, stocks, capital specificity, and supply-chain disruptions. Formal derivations distinguish marginal, average, fixed, variable, avoidable, sunk, and opportunity costs; worked examples establish the roles of constraints and time horizons. The research identifies the assumptions underlying upward-sloping competitive supply, shutdown thresholds, entry conditions, and long-run equilibrium. Comparative applications to agricultural commodities, manufacturing, infrastructure, and services clarify how cost concepts must be matched to feasible decisions and measurement regimes. The original analytical synthesis organizes these established results around the decision-specific conversion of resource sacrifices into market supply. This is a theoretical and selected-literature study, supported by historical texts, methodological documentation, and empirical research records; it introduces no original dataset or named theoretical system.

Keywords: production costs; supply formation; marginal cost; opportunity cost; capacity; competitive supply; entry and exit; cost pass-through; markups; production economics; economic theory; microeconomics

The scientific object and method

Production costs express the economic sacrifice associated with making goods and services available under specified technological, organizational, institutional, and temporal conditions. Supply formation is the process through which feasible production, resource commitments, and anticipated returns become quantities offered in a market. The decisive analytical question is therefore how a resource valuation becomes an output decision. A cost account records or imputes expenditures and sacrifices. A supply relationship describes choices conditional on prices and other determinants. A market equilibrium determines a compatible configuration of quantities and prices. These are three different objects, linked by explicit behavioral and institutional assumptions.

The unit of analysis matters. A plant can expand operating hours while the firm leaves its total capacity unchanged. A firm can transfer output among plants while industry output remains fixed. An industry can grow through entry while each incumbent operates at its previous scale. International supply can expand through imports while domestic production stagnates. A valid explanation identifies the decision maker, product, location, accounting boundary, time horizon, and constraints before comparing changes in costs or output.

This dossier treats the historical development of price theory, reconstructs standard cost and profit functions, and evaluates their scope through counterexamples and selected empirical studies. Sources were selected for direct relevance to the mechanism, not through a systematic search protocol or meta-analysis. Historical texts and accessible technical sections establish definitions and interpretations; selected empirical findings are reported within the scope of the available abstract or methodological material. No independent estimation, new dataset, or claim of theoretical priority is made. Formal examples use hypothetical monetary units and physical output units; all equations are analytical derivations under stated assumptions. The research belongs to Economic Theory and Microeconomics and concentrates on costs as determinants of supply and price, leaving a comprehensive treatment of production duality, game theory, and general equilibrium to separate research dossiers.

From classical price theory to marginal analysis

Smith's natural and market price (1776) provides an early account of the reproduction of production through remuneration of land, labor, and stock and of deviations caused by the relation between market quantities and demand backed by purchasing capacity. His natural price is historically specific: it presupposes ordinary compensation rates at a particular time and place. Market prices can depart from that reference due to scarcity, perishability, information, or restrictions on competition. This makes reproduction and adjustment central to the classical question even before a differentiable cost schedule is defined.

Ricardo's chapter on value (1817; 1821) adds the problem of relative labor requirements and the effects of capital composition and durability. His natural and market price chapter addresses the allocation of capital among employments as returns diverge. The classical project asks which conditions reproduce a system of production and distribution across activities. A contemporary supply function instead fixes some conditions provisionally and asks how a producer chooses an output level when a price varies. The two research objects intersect where producers require sufficiently attractive returns to retain and replenish productive resources, but neither the classical normal price nor labor-embodied value is simply the derivative of a modern cost function.

Marshall's analysis of long and short periods (1890; 1920) makes the time horizon an explicit determinant of normal supply conditions. Production facilities, specialized skills, and investment plans can adjust over longer intervals. This historical insight motivates a precise distinction between available capacity today and the resources required to reproduce capacity later. It also provides a reason to regard a given cost schedule as a conditional description rather than an eternal technical property.

Three questions emerge from this history. Which costs are incurred by an additional unit at an existing facility? Which revenues justify an additional facility or variety? Which distributional and institutional conditions reproduce the productive system? These questions require marginal, average, and intertemporal analyses respectively. The history of the subject supplies the questions and the subsequent formal apparatus supplies conditional answers. A comparison of theories should preserve both their explanatory level and the terms in which their original authors formulated their problems.

Economic cost and the boundaries of the decision

Explicit expenditures include wages, electricity bills, purchased intermediates, and contractual rents. Economic cost also includes the opportunity cost of owner-supplied labor, retained capital, and alternative uses of productive assets. The correct counterfactual is the best feasible alternative forgone, measured over the same horizon and with consistent risk treatment. A facility purchased for 100,000 monetary units years ago has a historical acquisition cost of 100,000; its economic sacrifice if retained today depends on available alternatives, including resale or alternative production. The two magnitudes can diverge sharply.

Fixed cost is constant with respect to the specified output decision within a stated range. Variable cost changes with it. Avoidable cost disappears if a specified action is not undertaken. Sunk cost is unrecoverable at the time of decision. These distinctions cross-cut one another. A fixed monthly licensing charge may be avoidable if renewal has not occurred. A quantity-dependent expense already incurred in preparing an order can be sunk at the moment the order is abandoned. A lease may be fixed but avoidable through a termination payment or subletting. The relevant cost difference must include the actual terms of those options.

The standard short-run cost definitions in OpenStax (2017) separate total cost C(q), fixed cost F, variable cost V(q), average cost C(q)/q, and marginal cost C'(q). With V(0)=0 and output q>0, average variable cost is V(q)/q; average fixed cost is F/q. The units of C are monetary amounts per decision period. Average and marginal costs are monetary amounts per unit. If output is measured in tons per week, both the production data and expenditure interval must refer to that week. A marginal change in energy use priced at a quarterly average tariff may misrepresent the actual marginal purchase when the tariff depends on time of day.

A resource acquired under a take-or-pay contract illustrates the central distinction. Using already committed inputs can have a low incremental cash expense, but the economic cost includes foregone resale revenue or another profitable use if such options exist. If no alternative is feasible and additional use requires no other resources, the opportunity cost of that input at the margin can be zero. The same contract may be costly in deciding whether to renew it. The evaluation changes because the available actions change, while the accounting payment itself can remain identical.

For private supply analysis the relevant prices are those faced by producers, including taxes and subsidies. Social opportunity cost additionally includes harms or benefits outside the producer's accounts. A decision that is profitable at prevailing prices can therefore consume resources with higher social value than its output. The appropriate cost depends on the question being asked: private behavior, financial reporting, resource allocation, or welfare. Their measurements should remain explicit rather than being combined into an ambiguous total.

Technology, input prices, and cost minimization

Let T(q) denote all nonnegative input vectors x capable of delivering at least q units of a homogeneous output, with technology and environmental conditions held fixed. For strictly positive input prices w, define the conditional cost function

C(w,q) = inf {w·x : x belongs to T(q)}.

When the set is nonempty and closed and the objective has bounded sublevel sets on it, the infimum is attained. The function is nondecreasing in q if higher-output requirements imply nested feasible sets, nondecreasing in each input price, positively homogeneous of degree one in w, and concave in w. Concavity follows because the pointwise infimum of linear functions is concave: each feasible x has a linear cost in input prices, and choosing the cheapest vector can moderate the effects of a relative price change. Concavity in input prices does not imply concavity in output. Output convexity depends on productive technology and the treatment of fixed costs.

Shephard's Theory of Cost and Production Functions (1971) is a foundational account of the relationship between technological representations and minimum-cost problems. For differentiable conditional cost with an attained optimum, the derivative with respect to an input price gives the corresponding conditional input demand. To see why, compare the minimized cost at two nearby vectors of prices using each optimum as a feasible alternative at the other price. These bounds approach the expenditure change based on the optimal input quantity when the relevant derivative exists. Nonunique minimizing bundles can yield a set of supporting input vectors instead of a unique derivative.

Consider technology q=A L^α K^β with A, α, β>0, and positive wage w and capital rental r. In a short period with K fixed at K̄>0, the required labor is L=(q/(A K̄^β))^(1/α). Variable labor cost is wL. If α<1, this cost is strictly convex in positive q and marginal cost rises with output. With both inputs adjustable, an interior minimum equates αq/(wL) and βq/(rK), or wL/(rK)=α/β. Defining s=α+β, the minimized cost scales as q^(1/s). Thus increasing returns to scale (s>1) give falling average cost in this stylized technology; decreasing returns (s<1) give rising average cost. A single-factor diminishing marginal product at fixed K can coexist with increasing long-run returns to joint input expansion.

In an alternative fixed-proportions technology q=min(L/a,K/b), a,b>0, the least-cost inputs are L=aq and K=bq, so C=(wa+rb)q when both factors can adjust. Relative costs alter unit expense, but conditional input proportions stay unchanged. If K is fixed, the quantity constraint q≤K̄/b can bind regardless of willingness to pay. Substitution is thus an empirical property of the technology and horizon, not an implication of a price increase alone.

Input substitution may require retooling, worker training, certification, financing, or coordination with suppliers. OpenStax (2017) describes the long-run cost adjustment associated with adjustable inputs. The analytical contribution of the comparison is to separate the cost of operating an established input bundle, the cost of switching bundles, and the value of additional flexibility. A producer may choose an expensive input today because delivery is reliable and alternative inputs are incompatible with installed facilities; the cheapest static bundle can be infeasible within the delivery horizon.

Marginal cost and competitive output

A price-taking producer chooses q in a feasible set Q to maximize π(q)=pq−C(q). Assume Q=[0,∞), C is continuous, and profit attains a maximum, as when cost grows faster than revenue at sufficiently large output. At a differentiable interior optimum, p=C'(q). If C is strictly convex, the objective is strictly concave and an interior solution is the unique global optimum. At q=0, a one-sided derivative condition may determine the corner; at finite capacity q̄, a binding upper constraint can determine output even when p exceeds marginal operating cost. With multiple local optima, comparing their objective values and the boundaries is necessary.

The monotonicity of competitive supply does not require differentiability or convexity. Suppose q₁ maximizes p₁q−C(q), and q₂ maximizes p₂q−C(q) at a higher p₂, with all other arguments and the feasible set unchanged. Optimality gives p₁q₁−C(q₁)≥p₁q₂−C(q₂) and p₂q₂−C(q₂)≥p₂q₁−C(q₁). Adding yields (p₂−p₁)(q₂−q₁)≥0, hence q₂≥q₁. At equal prices several outputs may maximize profit. This is a conditional ordering of global maximizers, not a claim that observed price and quantity necessarily move together.

At a regular interior point with C''(q)>0, dq/dp=1/C''(q), and the point elasticity is p/(qC''(q)). As cost curvature falls, the response rises locally. At a binding capacity limit output can remain unchanged; at discontinuous choices a derivative can fail to exist. Changes in costs, capacity, expectations, and demand regimes invalidate a comparison that treats everything but price as fixed. These qualifications identify the exact content of the law of supply as a comparative-static statement.

The textbook treatment of competitive output decisions (OpenStax, 2017) provides the familiar relationship between price, marginal cost, loss and operating choice. Its familiar graphical rule should be read as the solution of a maximization problem with the appropriate cost boundary. The global optimality argument above remains valid even for a producer with nonconvex production possibilities, so long as the feasible set and costs are unchanged and global maximizers exist. Nonconvexity can yield jumps and disconnected supply quantities; a smooth increasing curve is a special case rather than the general definition.

Shutdown, entry, and the economics of commitment

Suppose F is fully unavoidable for the current period and V(0)=0. Operating yields pq−F−V(q), while shutdown yields −F. A positive output is advantageous precisely when pq−V(q)>0 for some feasible q>0. If the infimum of average variable cost is attained, the minimum AVC gives the shutdown threshold under these assumptions. At equality, a positive operating quantity may tie with shutdown. Where fixed operating cost is avoidable, it must enter the comparison even though it does not enter the derivative for q>0.

For a hypothetical producer, let C(q)=100+4q+q²/2, q≥0, with the 100 incurred whether or not production occurs. Competitive supply is q=max(0,p−4). At p=10, q=6, variable cost=42, revenue=60, and economic profit=−82. Shutdown loses 100, so continued use saves 18. Average cost at q=6 is 142/6≈23.67 monetary units per unit. If 100 can instead be entirely avoided through shutdown, the producer chooses between zero and the same positive-output candidate with profit −82. It shuts down. Conditional on producing, the marginal cost remains 4+q in both cases; the participation decision changes because the cost at q=0 changes.

In this example the minimum average cost for positive production occurs at q=√200≈14.142, and the corresponding price is 4+√200≈18.142. If the whole fixed obligation can be avoided before entry, profit at the positive optimum is (p−4)²/2−100 for p>4. Entry is profitable above this threshold. The short-run operating threshold of 4 is therefore distinct from the full-cost recovery threshold. Real producers also face restart costs, contractual penalties, alternative use values, and future expected profits, so the same arithmetic must be extended to their actual counterfactual options.

An equipment purchase should be evaluated before the commitment using expected discounted benefits and costs. After commitment, the portion recoverable through resale or transfer remains relevant. The empirical study of displaced aerospace capital by Ramey and Shapiro (2001) uses equipment-level evidence from plant closures and reports discounted resale relative to replacement costs, particularly for specialized assets. The observation supports caution about equating book values or replacement prices with the opportunity cost of an installed asset. It does not establish a universal resale discount for other assets or industries.

Entry, operation, temporary shutdown, permanent exit, and later reentry are separate decisions. A firm can be profitable in continuing an operation while entry into an otherwise identical project would be unattractive. Existing productive capacity embodies the history of investment. Prices necessary to mobilize new productive resources may consequently exceed those sufficient to keep existing resources operating. This asymmetry is a direct implication of different feasible alternatives and avoidable costs.

Capacity, scale, and indivisibility

A short-run capacity restriction q≤q̄ makes production a constrained choice. If an interior marginal cost rule would give output above q̄, the producer supplies q̄. The shadow value of relaxing an active upper constraint is p−MC(q̄) under differentiability and the other assumptions of the price-taking problem. It measures the value of an additional unit of permitted output during the stated interval. It is not automatically the value of buying capacity: a capital investment adds output opportunities in future periods and has its own installation, financing, maintenance, and disposal costs.

In the quadratic example above, let q̄=8. At p=16, unconstrained supply would be 12, but output is 8. The marginal operating cost at the capacity limit is 12, so the local shadow value is 4 per additional unit of allowed output. If another facility can instantly and cheaply produce the remaining units, the constraint may disappear at the firm level while remaining binding at the original plant. A capacity estimate therefore requires a level of organization and time interval.

Indivisibilities give another form of cost behavior. A producer may require one oven for any positive production up to 100 units, then a second oven. If each oven entails a fixed avoidable charge F and marginal operating cost is c, a simple cost representation for q>0 is c q+F ceil(q/100). The marginal decision at 100 units involves a discrete increase, so the smooth derivative c alone cannot describe expansion beyond that point. At price p>c, a full additional block is profitable if 100(p−c)>F, assuming demand can absorb the block at p. A firm facing downward-sloping demand must instead compare additional revenue over the block with its full incremental expense.

The aggregation of heterogeneous indivisible decisions can create a relatively smooth industry relationship when many enterprises have distinct thresholds. It can also preserve discontinuities in concentrated industries with large facilities. Physical capacity, technical utilization, economic utilization, regulatory permissions, and available labor each describe different restrictions. A facility's maximum engineering throughput may exceed its feasible economic output if overtime premiums, quality defects, or delivery limits make continued expansion costly.

Increasing returns represent proportional input expansion generating more than proportional output. Economies of scale refer to falling minimized average cost. They coincide under particular assumptions, including comparable input prices and suitable technology, but observed decreases in average cost can also arise from spreading a fixed expense, changing input prices, improving process control, or changing product composition. Cost comparisons must distinguish these mechanisms to explain which response can recur when other producers expand.

Long-run supply and the reproduction of capacity

Long-run cost can be written C_L(q)=inf_K [F(K)+V(q;K)], where K denotes the adjustable capacity choice and F(K) is its economically relevant periodic cost. The cost envelope is no greater than the cost of any particular feasible fixed-capacity configuration evaluated with the same accounting convention. A differentiable interior optimum over K has zero derivative of total cost with respect to K, but a discrete or constrained choice requires global comparison. The envelope need not be smooth, and its marginal slope is meaningful only where it exists.

With identical potential entrants, free entry and exit, competitive pricing, constant input prices, a replicable technology, and an attainable finite minimum of long-run average cost, the zero-economic-profit benchmark satisfies p=min_q>0 C_L(q)/q. At an interior differentiable minimizing scale, MC=AC. Output per plant is then determined by the cost minimum while the number of plants adjusts to market demand, treating plant numbers as sufficiently divisible or the market as large. OpenStax (2017) discusses long-run entry and exit within this benchmark. Zero economic profit includes the required return on resources; it does not imply zero accounting profit or zero wages.

The industry-level result changes when expansion raises the price of scarce inputs, when local knowledge lowers costs as the industry develops, or when firms possess persistent heterogeneous capabilities. Let industry output be Q, and suppose an entrant's long-run unit cost is u(Q), incorporating industry scale effects. An equilibrium condition p=u(Q) can give rising, flat, or falling long-run equilibrium price with industry expansion depending on u'(Q), provided entry, demand and stability conditions are satisfied. A falling equilibrium locus does not contradict monotonic conditional individual supply because input prices or external economies vary across those equilibria.

Heterogeneous producers illustrate the distinction between an entrant's required return and incumbents' realized rents. If potential entrants differ in unit costs and access to an input is scarce, the marginal entrant may earn zero net economic profit while more efficient incumbents earn rents. Land, licensing rights, unique sites, or accumulated capabilities can have opportunity values that capitalize part of the rent. Whether an accounting estimate treats these resources as costly to the firm affects observed profitability but not the underlying allocation problem.

The transition between short and long run is a sequence of adjustments rather than one instant. Orders can trigger overtime within days, inventories can be replenished over weeks, training and procurement may take months, and facility expansion can take years. These horizons vary across activities. Long-run comparative statics identifies eventual configurations under assumptions; it does not predict the dates on which projects, firms, or industries reach them.

Industry supply, market clearing, and cost incidence

For n price-taking firms with supplies q_i(p;z_i), where z_i collects cost determinants, capacity and other state variables, market supply is Q_S(p;z)=Σ_i q_i(p;z_i). With competitive homogeneous output, an equilibrium solves Q_S(p;z)=Q_D(p;d), provided the feasible supply and demand values intersect. If supply is set-valued, the selected quantity must be compatible with each producer's optimization and market clearing. A fixed product price is taken as given for an individual decision; it is endogenous when determining market equilibrium.

For a hypothetical competitive market consider inverse demand p_D=100−Q and inverse supply p_S=20+Q. Solving yields Q=40 and p=60. A uniform per-unit cost increase of 12 shifts inverse supply to 32+Q. The new quantity is 34 and price paid by buyers is 66. Net receipts after the cost wedge are 54, compared with 60 previously. The cost wedge equals the sum of the 6-unit increase in buyer price and 6-unit reduction in producer net receipts. The result follows from equal absolute slopes and is not an empirical estimate.

More generally, for differentiable demand D(p_b) and supply S(p_s), with p_b=p_s+t and a uniform per-unit wedge t, differentiation gives dp_b/dt=S'/(S'−D'), where S'>0 and D'<0. Increasing the responsiveness of supply raises the buyer share in this local expression; more elastic demand reduces it. With capacity fixed, supply is locally vertical and the buyer price may remain anchored by demand for the fixed quantity while net receipts bear the wedge. If the wedge makes capacity unprofitable, this local result ceases to apply.

A producer's costs can rise because inputs become scarce, productivity falls, regulatory constraints tighten, or a currency movement raises import prices. These mechanisms have distinct consequences. Some shift marginal cost over all quantities; others lower maximum feasible output; others alter fixed entry costs without changing current operating cost. An input-price index is informative about one source of expenditure pressure but is insufficient to infer the shape or magnitude of aggregate supply adjustment.

Trade introduces further conditions. With unrestricted homogeneous imports at a given delivered world price, domestic price may be tied to that price while domestic production contracts and import volume rises. Transport, tariffs, capacity constraints at ports, exchange-rate movements and product differentiation weaken that limiting case. Domestic cost shocks need not have the same price consequences in small open economies and in markets where domestic production sets the marginal source of supply.

Market power, setup cost, and the limits of a supply curve

When an enterprise recognizes that its chosen quantity affects the price of its product, revenue becomes R(q)=P(q)q. A differentiable interior profit maximum satisfies P(q)+qP'(q)=MC(q), with global and second-order conditions still required. Price and output depend jointly on demand and cost. The familiar competitive supply curve is generally absent for a monopolist because the same market price can be associated with different chosen quantities under different demand schedules.

Consider two independent hypothetical markets with the same monopolist marginal cost c=10. Market A has P_A(q)=50−q. Market B has P_B(q)=50−2q. In A, q=20 and p=30; in B, q=10 and p=30. Linear demand and constant marginal cost ensure strictly concave profit. The same seller price and marginal cost produce different quantities because demand slopes differ. Assigning a single quantity to price alone would erase the strategic determination of revenue.

With fixed setup cost and constant marginal expense C(q)=F+cq for q>0, average cost exceeds c at finite output. Pricing every unit at marginal cost leaves F uncovered if no separate payment or subsidy exists. Dixit and Stiglitz (1977) develop this tension in Monopolistic Competition and Optimum Product Diversity, where scale economies and the value of variety affect welfare and the number of products. Their results depend on demand structure and the assumed costs; the existence of setup costs alone does not establish whether entry is excessive or insufficient.

In a simple symmetric constant-elasticity example with demand elasticity magnitude σ>1, and with each producer treating the relevant demand shifter as fixed, an interior price satisfies p=cσ/(σ−1). With free entry and zero economic profit, (p−c)q=F implies q=F(σ−1)/c for c>0. If F=120, c=6, and σ=3, this gives p=9 and q=40; revenue 360 equals variable cost 240 plus fixed cost 120. The example shows how a price above marginal cost can coexist with zero economic profit. The number of varieties and equilibrium demand shifter require additional market-clearing conditions.

For joint products, cost recovery is more complex still. A single process may yield output A and output B in fixed proportions. The opportunity cost of producing additional A depends on the value or disposal expense of B. An allocated accounting expense assigned to A can help internal reporting but is not necessarily the incremental cost of A. Pricing can depend on the entire portfolio of demand, process constraints, and alternatives. Similarly, networks with common infrastructure may recover common costs through access charges and usage fees. Welfare analysis must separately evaluate participation, consumption, investment, and the distribution of common-cost recovery.

Cost shocks and price pass-through

Pass-through measures the response of an output price to a specified cost change. It is essential to state whether it is a ratio of monetary changes (dp/dc), an elasticity (d ln p/d ln c), or a response measured over a time interval. When the final price includes inputs unaffected by the shock, a low elasticity with respect to one component can coexist with substantial transmission of changes in total marginal cost.

Under a monopolist with inverse demand P(q)=a−bq, b>0 and constant marginal cost c<a, the solution is q=(a−c)/(2b), p=(a+c)/2, so dp/dc=1/2. Under constant-elasticity demand q=A p^(−σ), σ>1, the unconstrained interior price at constant marginal cost c>0 is p=cσ/(σ−1), giving monetary pass-through σ/(σ−1)>1 but elasticity exactly one. Both are profit-maximizing examples. The difference arises from demand curvature, not from a change in the magnitude of the production cost shock. In imperfect competition a universal one-for-one price response is therefore untenable.

Weyl and Fabinger (2013) extend tax incidence reasoning in Pass-Through as an Economic Tool to imperfect competition. For empirical analysis a cost shock must be distinguished from coincident changes in consumer demand, input quality, contractual arrangements, and competitor behavior. Where contracts fix retail prices until renewal, contemporaneous price changes can understate cumulative transmission. Inventory valuation and replacement costs can create different recorded input expenses for enterprises facing identical current market prices.

Nakamura and Zerom's Accounting for Incomplete Pass-Through (2010) studies coffee prices with a structural oligopoly framework. Its published abstract reports that local costs and markup variation reduce estimated long-run pass-through relative to a specified benchmark and that adjustment frictions primarily account for delay. The evidence concerns the industry and model examined; it does not imply that a fixed percentage of commodity shocks reaches consumers everywhere. The wholesale and retail stages can have distinct response schedules, so an aggregate retail-price series alone does not establish where the adjusting margin lies.

A research design should use an explicitly measured exogenous cost component, estimate responses over a predeclared horizon, compare similarly exposed and less exposed products or firms, and verify changes in quantities and markups when data permit. Where the shock affects input availability as well as price, a single pass-through parameter cannot identify both cost shifts and rationing without further information. The analyst should preserve evidence of delayed adjustments, temporary margins, promotions and changes in product quality rather than forcing the data into an instantaneous price equation.

Irreversibility, uncertainty, and time

A multi-period decision chooses current output, capacity additions, and possibly shutdown or expansion subject to state transitions. A minimal expression is the maximum over feasible actions of E Σ_t δ^t [p_t q_t−V_t(q_t,K_t)−I_t], where I_t denotes investment expenditures, δ is a discount factor between zero and one, and K evolves with investment and depreciation. This is an illustrative standard discounted formulation, not a new authorial model. Operating expenses and investment must be recorded without double-counting capital charges. A terminal asset value, debt obligations, taxes and adjustment costs can be added consistently when relevant.

Uncertainty changes the meaning of supply responsiveness when investments are hard to reverse. Waiting can preserve information and alternative choices. A firm need not expand at the first observation of price exceeding the estimated current operating cost. If investment is reversible and markets complete, some of these considerations diminish; when equipment is specialized and reentry costly, the comparison is asymmetric. The asset-specific resale evidence discussed earlier explains why the degree of reversibility cannot simply be assumed from an engineering description of equipment.

Inventories can separate production from sales. If a firm's stock balance is B_(t+1)=B_t+q_t−s_t−loss_t, current deliveries s_t may increase without current output q_t increasing. A high expected future price can also induce a seller to hold stocks back from the present market. For perishable products, storage losses and shelf life constrain this margin. A production supply curve and a release-from-inventory decision are consequently distinct. A market-price response to anticipated scarcity can precede any observed change in the factory's production costs.

Credit and working capital can limit input purchases before revenues arrive. Suppose a producer has liquid resources M and must pay variable expenditures in advance. If operating cost V(q) must satisfy V(q)≤M, a profitable unconstrained output may be inaccessible. An additional unit of liquid funds has a shadow value when this constraint binds. Importantly, a high financing cost affects desired output differently from a hard credit ceiling. Observing an unchanged plant size with declining production can therefore indicate reduced input access, a higher marginal price, or demand weakness. Distinguishing them requires direct financial and procurement evidence.

The timing of investment also shapes measured productivity. A plant preparing for new capacity may temporarily reduce output while incurring installation costs. A mature plant may show higher short-period output by postponing maintenance at the expense of future reliability. Intertemporal cost evaluation must compare sustainable output paths, not only current measured expenditures. The opportunity cost of capacity is expressed through the alternatives over its entire relevant use horizon.

Cost measurement and the evidence available to researchers

A credible empirical cost study begins with reconciliation of physical production records, product mix, hours, intermediate quantities and prices, inventories, and capital services. The object to be explained is a counterfactual resource expense for a change in feasible output. Accounts typically provide realized transactions and imputations. Deriving marginal costs from them requires variation and a model of what changes with output. Dividing total expenditures by sales yields a ratio; differentiating an estimated cost function yields a conditional marginal cost. Their economic meanings are distinct.

The OECD (2009) Measuring Capital manual organizes the measurement of asset stocks, depreciation and capital's productive role. A stock value is measured at a date; a flow of productive services pertains to a period. In an illustrative single-period rental calculation, an asset of opening value V that depreciates at rate d and carries required return r has a real user cost approximately (r+d−g)V when expected real asset appreciation is g and the chosen simplifications omit taxes and higher-order terms. The precise expression depends on timing, asset survival, financing and valuation conventions. A historical acquisition expenditure cannot simultaneously be charged in full as a current operating cost and counted again through its services without reconciliation.

The BLS industry productivity concepts distinguish output relative to labor and output relative to combined inputs. The analyst should likewise separate changes in output price from changes in output quantity, hours from head count, and gross output from value added. A common unit labor cost measure is compensation divided by real output, or hourly compensation divided by real output per hour when hours concepts align. It identifies labor cost pressure per output unit, not total marginal cost. Intermediate purchases, capital, quality, utilization, and taxes may change differently.

In a manufacturing example, annual expenditure on labor is 200,000, intermediates 300,000, economic capital services 100,000, and physical output 10,000 identical units. Average total economic cost is 60. If output rises to 12,000 and expenditure rises to 670,000 while comparability is maintained, average cost falls to about 55.83 and the interval incremental cost is 35. This interval ratio is neither the point derivative nor proof of a permanent technological change. Output composition, input prices, overtime, maintenance timing and stock accounting might explain part of the difference.

Production measured by revenue divided by an industry price index can confound firm-specific prices and physical efficiency. Foster, Haltiwanger, and Syverson (2008), in Reallocation, Firm Turnover, and Efficiency, use separately observed quantities and producer prices to investigate this distinction. The implication is methodological: a firm that raises its revenue through stronger demand need not have become more technically productive, while a more efficient firm can lower its price. A researcher should identify whether the productivity measure includes market valuation before attributing residual variation to cost differences.

Estimating marginal cost, markups, and supply elasticity

The price-cost markup is μ=p/MC for positive marginal cost, while the Lerner index is (p−MC)/p=1−1/μ. A gross profit margin from accounts usually deducts a specified cost category from revenue. It can differ from both measures through fixed costs, changes in inventory, product mix and the exclusion or allocation of economic capital charges. Marginal cost is rarely observable directly for differentiated products.

Hall's price and marginal cost study (1988) uses annual variations in inputs and output to address price–marginal-cost differences across U.S. industries. The published abstract reports evidence of prices exceeding estimated marginal cost in some industries. Interpreting such variation requires assumptions about technological changes and other unobserved determinants of production. It does not yield a timeless sector average or establish that all firms charge positive economic rents.

For a firm choosing variable input x with price w, let f(x,k) be physical output and ε_x=(∂f/∂x)x/q be its output elasticity. Cost minimization at an interior solution gives w=MC·∂f/∂x. Dividing xw by pq then yields the revenue expenditure share s_x=ε_x/μ, so μ=ε_x/s_x. This derivation requires an appropriate price-taking input market, variable input choice, differentiable production, accurate physical or appropriately modeled output, and absent unmodeled wedges in this first-order condition. If a procurement quota binds or an input has adjustment frictions, the expression must be modified.

De Loecker and Warzynski (2012) use the production approach in Markups and Firm-Level Export Status; the publisher abstract describes the cost-minimization and variable-input assumptions and reports differences associated with export status. This is a model-based inference, not an observation of actual marginal-cost invoices. A major related identification problem is the correlation between firms' input decisions and unobserved productivity. Olley and Pakes (1996), in The Dynamics of Productivity, address simultaneity and selection in a setting with entry, exit and changing plant scale. The findings on telecommunications equipment do not establish the same reallocation process in unrelated sectors.

Supply elasticity presents a separate simultaneous-equations problem. In an observed market p and Q are determined by the intersection of demand and supply. Regressing Q on p combines movements along and shifts of both relationships. If an independent demand shock changes price while leaving the supply schedule unchanged, it can identify a supply response under exclusion, relevance, and the specified dynamic conditions. If the shock also moves input costs or capacity, the exclusion assumption fails. Instrumental variables, structural models and natural experiments provide alternative strategies whose validity depends on their underlying causal assumptions.

An empirical program should report output definition, unit of observation, data window, treatment of entry and exit, identification strategy, instrument or model diagnostics, sampling uncertainty, and sensitivity to price deflation and input treatment. It should distinguish a short-run within-plant elasticity from long-run industry expansion and from an equilibrium response involving imports. Comparison of estimates without those distinctions can create apparent contradictions that merely reflect different estimands.

Agriculture and storable commodities

Agriculture makes the timing of supply particularly visible. Land allocation, planting, cultivation and harvest are successive decisions. Weather influences yield even if acreage has already been committed. Short-run output may be biologically constrained while next-season acreage and input intensity respond to expected returns. The relevant output price for planting can be a forward expectation rather than the current spot price. Land quality, irrigation, seed technology, input availability, and financing shape the feasible response.

USDA ERS Commodity Costs and Returns documentation describes a methodology based on periodic producer surveys and annual updating of price, production, and acreage information. The documentation identifies survey intervals of roughly four to ten years by commodity. This matters because the underlying technology is connected to the survey vintage. The published accounts are valuable for comparing cost categories and patterns within their methodology, but annual updates do not amount to independent full surveys each year. Comparisons across countries should align land tenure, unpaid labor, financing, yields, farm size, allocation of joint costs, and whether costs are per hectare or per harvested unit.

Roberts and Schlenker (2013) study agricultural supply and demand elasticities using past yield shocks and their effects on inventories and futures prices to help identify supply relationships for storable commodities. Their strategy illustrates how physical shocks and storage dynamics can create variation informative about price sensitivity. It also clarifies why a contemporaneous yield shock is unsuitable as an excluded demand shifter when estimating supply: it alters the feasible quantity itself. Estimates are conditional on the commodity aggregate, data period, model, and identifying assumptions.

A hypothetical farmer expects output price 200 per ton, yield five tons per hectare and avoidable preplanting cost 750 per hectare. Expected revenue is 1,000, giving 250 before any further relevant land, risk, and other costs. If competing use of the land yields 300 net per hectare under comparable risk and timing, planting is economically inferior despite positive cash contribution. After planting, some expenses are sunk, and harvest should be evaluated using avoidable harvest and marketing expenses, expected additional output revenue and alternative uses. This illustrates why observed marketing can continue after prices fall below a preplanting break-even estimate.

At a global level a drought can reduce harvest supply in one region and raise expected prices, encouraging production elsewhere in a later season. Trade and storage partly redistribute the shock across space and time. Global quantity response and the distribution of impacts depend on land constraints, crop substitution, tariffs, transport and inventory levels. A static domestic agricultural cost schedule alone cannot identify these interacting responses.

Manufacturing, supply chains, and critical inputs

Manufacturing supply combines transformation technology with procurement and logistics. Plants incur expenditures on equipment, energy, labor, materials, quality assurance, inventories, and downtime. A comparatively inexpensive component can be a critical bottleneck if production requires it in fixed proportion. Its expenditure share and its technical importance are different magnitudes. The feasible set can contract abruptly even while the price of other materials falls.

For a hypothetical process q=min(m/2,l/3), m denotes component units and l labor hours. Required proportions are two components and three hours per unit. With available m=100 and l=300, maximum feasible output is 50. Raising labor to 360 cannot raise output without additional components. If component availability falls to 80, the maximum becomes 40 despite unchanged wages and equipment. A smooth marginal cost function estimated during normal procurement can mispredict this situation because the binding state variable has changed.

Boehm, Flaaen, and Pandalai-Nayar (2015) investigate input linkages after the Tohoku earthquake in a Federal Reserve working paper. Its abstract reports sharp output declines among affected U.S. affiliates of Japanese firms, roughly proportional to import declines, and estimates consistent with low substitutability over the examined period. The setting supplies direct evidence for the possibility of cross-border bottlenecks. The finding should not be generalized to all suppliers, all industries, or permanently fixed coefficients. Reengineering and qualification of substitute suppliers may change the long-run response.

When a manufacturer faces a sharp input-price increase, measured average cost may rise because existing inputs are used more expensively or because output falls and fixed expense is spread over fewer units. The latter can occur even with unchanged marginal prices. At low utilization, average cost can therefore be high while the expenditure required to produce an additional unit remains modest. A firm may continue production, hold idle capacity, change suppliers, or abandon a product depending on avoidability, future demand, and the expected cost of restarting.

A supply chain links many such decisions. Upstream shortages can alter the attainable input sets of downstream firms. Inventories, alternative suppliers, shared transport infrastructure and contractual priorities mediate the effect. International comparative studies should identify whether an observed cost reflects local production, imported inputs, landed cost or the delivered cost at the customer's location. A shipment's transaction price can include margins, logistics and trade charges that are not the resource cost of the underlying factory process.

Infrastructure, services, and digital production

In a capacity-constrained infrastructure market, several producers may offer units at differing incremental operating costs. A stylized dispatch problem minimizes Σ_i c_i q_i subject to Σ_i q_i=D, 0≤q_i≤K_i, identical delivered product and no other system constraints. Order the c_i from lowest to highest. The solution fills available lower-cost capacity first. The shadow value of another unit of demand equals the marginal available cost where the problem is regular; when all capacity is exhausted, an additional unit may be infeasible. In actual electricity systems, network limits, ramp rates, startup costs, minimum run times and settlement rules complicate this stylized example. It is an analytical illustration, not a description of a specific country's market rules.

Suppose two technologies have capacity 60 and 40 units, with marginal operating costs 20 and 50. At demand 80, least-cost dispatch uses 60 from the first and 20 from the second, total operating expense 2,200 and marginal operating cost 50. With demand 50, the first technology alone supplies it and marginal cost is 20. The 30-unit change in marginal cost arises from a change in the dispatched technology, even though each technology's cost remains constant. Recovering the infrastructure's fixed investment requires a separate evaluation of total revenues across time and contractual or regulatory arrangements.

Services have their own capacity dimensions: available appointments, personnel time, utilization of premises, and customer processing time. A vacant appointment can be relatively inexpensive to fill shortly before it would otherwise expire, while long-run expansion requires additional staffing and facilities. Quality can vary with utilization if waiting times, errors or service interruptions increase. A service quantity measure should therefore record quality and timing when they are economically material.

Digital products often combine substantial development expenditure with relatively small incremental replication expense. Additional users can nevertheless cause meaningful infrastructure, support, compliance, energy, and capacity costs. Near-zero incremental replication cost does not imply zero total marginal service cost at every scale or zero opportunity cost of data, computing resources and engineering effort. Subscription, bundled or access pricing may spread common costs across users, but whether a given arrangement is feasible or efficient depends on demand, congestion, competitive alternatives and cost structure.

These sector comparisons show why the relevant form of cost is tied to the actual production process. Agricultural yields, manufacturing bottlenecks, infrastructure dispatch and service capacity are four ways in which feasible supply can change. They should not be compressed into a single universal elasticity or assumed to share the same adjustment interval.

Private costs, public policy, and social opportunity costs

Private supply decisions respond to prices and constraints facing the producer. An emissions externality, for example, can impose resource losses on other parties outside the producer's accounts. In a differentiable partial-equilibrium model with private cost C(q) and external damage D(q), social marginal cost is C'(q)+D'(q). If demand measures marginal social benefit, the efficient interior quantity under the model satisfies marginal benefit equal to social marginal cost. Implementing that quantity through a per-unit charge equal to marginal external damage at the optimum requires additional institutional and informational conditions; distributional effects and administrative feasibility remain relevant.

A subsidy can lower the private price of expanding output without reducing the real resources necessary for its production. Conversely, a costly regulation can improve product safety or environmental quality, which may alter both the definition and value of output. Treating all compliance expenditure as pure social loss would overlook the benefits being purchased. Comparative analysis needs to identify the counterfactual quality level and the distribution of costs across producers, consumers and other affected parties.

A price ceiling below the clearing price can produce excess demand if offered quantities fall short of desired quantities at the controlled price. Under rationing, waiting time, eligibility, product quality, informal charges or delivery priority can become economically consequential. A price floor can lead to excess offered quantity and changes in entry or quality. These conditional results follow from the relationship between feasible supply, demand and price constraints; their empirical size requires evidence about implementation and behavior. In markets with market power or externalities the welfare effects of interventions require case-specific comparison.

The costs of entry and adjustment can also be shaped by public infrastructure, standards, training, licensing and market access. Reducing a bottleneck can create capacity even when the direct input price has not changed. International differences in institutional reliability affect the opportunity cost of working capital and inventory. The purpose of cost analysis is to identify the real margin influenced by an intervention; changes in private prices, measured expenses, physical production and welfare should be tracked separately.

Comparative research design and falsifiable implications

The cross-country research program should first define a product with comparable specifications, production stage and reference period. A producer price received at the factory gate and a retail price including taxes and distribution are different observations. Input prices must correspond to the amounts actually paid or sacrificed at the relevant margin; exchange rates, inflation adjustments and purchasing power comparisons answer different questions. Convert monetary values using a declared convention and retain physical quantities for productivity analysis.

Next identify the decision period and capacity state. Classify costs by both their output dependence and avoidability. Record the treatment of depreciation, capital services, unpaid labor, owned land and financing. A fixed-versus-variable classification suitable for one month may be unsuitable for five years. Report cost uncertainty and the sensitivity of estimated marginal costs to the accounting boundary.

Three specific propositions can guide research without presuming their empirical confirmation. Under stable costs and fixed capacity, competitive optimal output should be weakly nondecreasing in independently varied seller prices. This proposition can be evaluated using demand-driven price changes with credible exclusion restrictions; a negative measured relation may instead reflect simultaneous cost or capacity shifts. A positive price shock may have little immediate effect but a larger later effect when replanting, investment or entry becomes feasible. Testing the temporal proposition requires a panel with clearly identified horizons and attention to expectations. Finally, in a cost shock concentrated in one input, the output-price elasticity need not equal the percentage increase in that input's price. Testing this requires input expenditure shares, price-setting behavior and adjustment timing.

Additional tests address participation. When a cost becomes avoidable through stopping operations, firms with otherwise comparable expected contributions should be more likely to stop than those tied to unavoidable commitments, holding alternative opportunities and expected future margins fixed. This is an implication of the specified choice structure. It is not a claim that every firm follows the same rule in real time: information, credit, contracts, strategic behavior and uncertainty may alter the choice set.

A useful empirical comparison reports separately observed facts, estimated parameters, model assumptions, and derived counterfactuals. Natural experiments may improve identification but do not automatically establish worldwide generalizability. Structural estimates permit counterfactual simulation at the cost of assumptions about technology and behavior. Cost-accounting studies support measurement but do not alone establish causal responses. Triangulation should compare these methods while preserving what each can identify. The cited evidence spans historical and modern research, U.S. cost and productivity statistics, international input linkages and OECD measurement concepts; it is not a representative sample of every national economy.

Original analytical synthesis

The central synthesis is that production cost explains supply through the specific decision for which that cost is relevant. A marginal expense governs the intensity of production within capacity; an avoidable commitment governs participation; the expected economic cost of durable resources governs replenishment and expansion. The mathematical distinction is consequential because the same underlying expenditures can influence different decisions in different ways. This is an integrated interpretation of established economic analysis, without a claim to a new named theory or theorem.

The first consequence concerns a common interpretation of prices. A price below average total cost cannot alone establish that current production should stop. The avoidable cost of stopping must be compared with revenues and the value of alternatives. Equally, a price higher than marginal operating cost cannot alone establish that new capacity should be built. The costs of investment, future utilization and irreversibility belong to that decision. These two errors arise from applying the correct cost concept to the wrong choice.

The second consequence concerns the relation between technology and market structure. Cost minimization maps a technology and input prices into conditional costs. Competitive profit maximization maps these costs and an output price into a supply correspondence. Market clearing determines equilibrium output and price. If firms have strategic price influence, the final step requires demand and the competition mechanism inside every firm's decision. If plants share inputs or products, the relevant cost becomes joint or conditional. Consequently an empirical cost schedule can discipline market analysis without fully determining prices by itself.

The third consequence concerns measurement. A decline in average cost can coincide with an increase in marginal cost if fixed expenses are being spread while operating constraints tighten. A high markup can coexist with zero economic profit when fixed costs are recovered. A producer's revenue productivity can improve without physical productivity increasing. These are not paradoxes: the underlying quantities refer to different denominators, margins or counterfactuals. The interpretation of data should precede any comparison of levels or growth rates.

The fourth consequence concerns disruptions. A bottleneck can contract the feasible set while conventional price statistics change little. A replacement cost shock can raise the expected cost of replenishing inventories without immediately altering reported historical expenses. A nominal price movement can shift the location or timing of production rather than current global quantity. Cost analysis that incorporates physical feasibility and contractual commitment explains why the same observed price movement can have qualitatively different output consequences.

The fifth consequence is that the long run describes a different set of possible actions, not simply a more responsive version of today's factory. New enterprises, technologies, sites, contracts, and supporting institutions can emerge; asset liquidation and specialized skills may impede transition. The most useful comparison is therefore between specified state-contingent decisions and their feasible alternatives. This perspective integrates the classical concern with reproduction, the marginalist logic of choice, and contemporary empirical research on uncertainty, capacity, cost measurement and market structure. It supplies a disciplined basis for research on actual industries without claiming to replace those scientific traditions.

Conclusion

Production costs become determinants of supply when the producer's feasible choices, prices, resource alternatives and time horizon are specified. Cost minimization translates technology into a conditional expenditure function. Profit maximization translates that function into output and participation decisions. Competitive equilibrium connects individual responses with demand, while market power, common costs, capacity, and institutional restrictions require further structure.

The main results of this research are conditional and mutually consistent. A competitive producer's global optimal output is weakly increasing in price when its costs and opportunities are unchanged. The equality of price and marginal cost characterizes suitable interior decisions, not every decision. Shutdown depends on avoidability, while entry requires expected full economic cost recovery. Industry expansion depends on the reproduction of capacity and the availability and cost of inputs. The elasticity of supply, the slope of a cost schedule, and the pass-through of cost shocks measure distinct phenomena.

Historical price theory helps explain the temporal and reproductive dimensions of supply; contemporary formal and empirical research clarifies the decisions and measurements involved. The comparison reveals why accounts alone cannot determine marginal costs, why capital specificity and intermediate-input bottlenecks matter, and why market structure and demand govern the transmission of cost changes into final prices.

Further research can improve sector-specific measurements of avoidable costs, capacity and capital services; assess the causal effects of credit and procurement constraints; and compare the time required for supply adjustment across countries and production systems. The appropriate next step depends on the empirical object. Within its analytical scope, this dossier establishes the relation among resource sacrifice, feasible production, participation, market coordination and the intertemporal reproduction of supply.

References

Boehm, Christoph E., Aaron B. Flaaen, and Nitya Pandalai-Nayar (2015). Input Linkages and the Transmission of Shocks: Firm-Level Evidence from the 2011 Tohoku Earthquake. Finance and Economics Discussion Series 2015-094. Washington, DC: Federal Reserve Board. DOI: 10.17016/FEDS.2015.094. Verified record.

De Loecker, Jan, and Frederic Warzynski (2012). Markups and Firm-Level Export Status. American Economic Review 102(6): 2437–2471. DOI: 10.1257/aer.102.6.2437. Verified record.

Dixit, Avinash K., and Joseph E. Stiglitz (1977). Monopolistic Competition and Optimum Product Diversity. American Economic Review 67(3): 297–308. Journal scan hosted by TU Wien. Verified record.

Foster, Lucia, John Haltiwanger, and Chad Syverson (2008). Reallocation, Firm Turnover, and Efficiency: Selection on Productivity or Profitability? American Economic Review 98(1): 394–425. DOI: 10.1257/aer.98.1.394. Verified record.

Hall, Robert E. (1988). The Relation between Price and Marginal Cost in U.S. Industry. Journal of Political Economy 96(5): 921–947. DOI: 10.1086/261570. Verified record.

Marshall, Alfred (1890; 8th ed. 1920). Principles of Economics. Book V, Chapter V, Equilibrium of Normal Demand and Supply, Continued, with Reference to Long and Short Periods. London: Macmillan. Econlib, eighth-edition text. Verified record.

Nakamura, Emi, and Dawit Zerom (2010). Accounting for Incomplete Pass-Through. Review of Economic Studies 77(3): 1192–1230. DOI: 10.1111/j.1467-937X.2009.589.x. Verified record.

OECD (2009). Measuring Capital—OECD Manual 2009: Second Edition. Paris: OECD Publishing. 233 pp. DOI: 10.1787/9789264068476-en. Verified record.

Olley, G. Steven, and Ariel Pakes (1996). The Dynamics of Productivity in the Telecommunications Equipment Industry. Econometrica 64(6): 1263–1297. Verified record.

OpenStax (2017). Principles of Economics 2e. Section 7.3, Costs in the Short Run. Houston: Rice University, OpenStax. Verified record.

OpenStax (2017). Principles of Economics 2e. Section 7.5, Costs in the Long Run. Houston: Rice University, OpenStax. Verified record.

OpenStax (2017). Principles of Economics 2e. Section 8.2, How Perfectly Competitive Firms Make Output Decisions. Houston: Rice University, OpenStax. Verified record.

OpenStax (2017). Principles of Economics 2e. Section 8.3, Entry and Exit Decisions in the Long Run. Houston: Rice University, OpenStax. Verified record.

Ramey, Valerie A., and Matthew D. Shapiro (2001). Displaced Capital: A Study of Aerospace Plant Closings. Journal of Political Economy 109(5): 958–992. DOI: 10.1086/322828. Verified record.

Ricardo, David (1817; 3rd ed. 1821). On the Principles of Political Economy and Taxation. Chapter I, On Value. London: John Murray. Econlib, third-edition text. Verified record.

Ricardo, David (1817; 3rd ed. 1821). On the Principles of Political Economy and Taxation. Chapter IV, On Natural and Market Price. London: John Murray. Econlib, third-edition text. Verified record.

Roberts, Michael J., and Wolfram Schlenker (2013). Identifying Supply and Demand Elasticities of Agricultural Commodities: Implications for the US Ethanol Mandate. American Economic Review 103(6): 2265–2295. DOI: 10.1257/aer.103.6.2265. Verified record.

Shephard, Ronald W. (1971). Theory of Cost and Production Functions. Princeton, NJ: Princeton University Press. Princeton Legacy Library reissue, 2016. ISBN 9780691647524. Verified record.

Smith, Adam (1776). An Inquiry into the Nature and Causes of the Wealth of Nations. Book I, Chapter VII, Of the Natural and Market Price of Commodities. London: W. Strahan and T. Cadell. Econlib, Cannan edition. Verified record.

U.S. Bureau of Labor Statistics (n.d.). Handbook of Methods: Industry Productivity, Concepts. Washington, DC: BLS. Accessed November 17, 2026. Verified record.

U.S. Department of Agriculture, Economic Research Service (n.d.). Commodity Costs and Returns: Documentation. Washington, DC: USDA ERS. Accessed November 17, 2026. Verified record.

Weyl, E. Glen, and Michal Fabinger (2013). Pass-Through as an Economic Tool: Principles of Incidence under Imperfect Competition. Journal of Political Economy 121(3): 528–583. DOI: 10.1086/670401. Verified record.